Skip to main content

Full text of "Ethnology"

See other formats


I I v 

co>- si 

nX Tl <r* 


!sj ^g 

Cambridge <grap!)ual 






Berlin: A. ASHER AND CO. 
Urijjjig: F. A. DROCKHAUS 
fUto $ork; G. P. PUTNAM'S SONS 
anU ffalrufta: MACMILLAN AND Co., LTD. 

All rights reserved 





A. H. KEANE, LLD., F.R.G.S. 




First Edition, 1895 
Second Edition, iN<X> 
Reprinted 1901, 1909 





COMPREHENSIVE English works on ETHNOLOGY in the stricter 
sense of the term, works such as those of Dr Prichard, Messrs 
Nott and Gliddon, and Dr Latham, were all composed before 
the appearance of Darwin's Origin of Species (1859), with 
which biological studies make a fresh start. Since then special 
branches of the subject, such as the evolution and antiquity ot 
Man, primitive culture, the Stone and Bronze Ages, and the 
origin of civilisation, have been treated by several eminent men 
of science, conspicuous among whom are Sir Charles Lycll, 
Professor Huxley, Darwin himself, Professor E. B. Tylor, Sir 
John Evans, Sir John Lubbock and Professor Boyd Dawkins. 
But scarcely any serious work of a comprehensive character 
can be mentioned except Dr Brinton's Races and Peoples, and 
Professor Tylor's popular treatise on Anthropology \ which, despite 
its title, is concerned more with ethnological and social than with 
strictly anthropological matters. 

When, however, the foreign literature of the subject is taken 
into account, a literature enriched by such eminent names as 
those of Virchow, Bastian, Waitz, and Kollmann in Germany ; 
Retzius, Castr^n, Worsaae, Forchhammer, Steenstrup and Mon- 
telius in Scandinavia ; Broca, Topinard, de Quatrefages and 
Hamy in France; Sergi, Mantegazza and Giglioli in Italy; it 
becomes evident that, since the general acceptance of evolutionary 
teachings, sufficient materials have already been accumulated to 
justify M. de Lapparent's declaration that Pheure des granites syn- 
theses a d'cja sonne. Such a synthesis is here for the first time 

viii PREFACE. 

attempted in the English language, in the hope that, even if 
but partly successful, it may still be accepted as a boon by those 
students who acutely feel the want of some trustworthy guide, 
especially amid the initial entanglements of a confessedly difficult 
subject. A work speaking with uncertain sound would obviously 
be useless, or at least of little value, for this purpose. Hence what 
might otherwise be regarded as a somewhat dogmatic treatment 
is here necessarily adopted, even in respect of many perplexing 
problems which till lately might justly be regarded as moot ques- 
tion^ on which it would be rash to pronounce a definite opinion 
either way. But for those who frankly accept its essential principles 
evolution is found to be a golden " skeleton key," which readily 
opens the door to many secret chambers, even in the more 
recondite recesses of human knowledge. Take, for instance, the 
origin of articulate speech, a question which in pre-Darwinian 
times was necessarily relegated by naturalists to the region of pure 
metaphysics, and which by anti-evolutionists is still regarded 
either as insoluble, or as soluble only by the assumption of direct 
creative force. Now, however, it is easily seen by anthropologists 
that language, like man himself, had a very humble beginning, 
and has reached its present marvellously perfect state sensim sine 
sensu, slowly improving in its phonesis and structure hand in 
hand with the slow improvement of the physical organs in virtue 
of which man has become a speaking animal. Its inner me- 
chanism is analysed by the comparative philologist, and found 
to be reducible to simple elements, and this conclusion is con- 
firmed by the comparative anatomist, who points out with Dr 
Arthur Keith that the facial organs of speech are non-existent in 
the anthropoids, rudely developed in fossil man, and perfected 
only in later ages. Thus is revealed the origin of language, which 
does not drop ready-made from the skies, but grows up from 
crude beginnings on the earth. The sources of much false 
reasoning and mystification are thus removed, and the truth 
stands out plain enough for all those willing to accept it. 

From this view* of its origin there directly follow other 
important inferences regarding the nature and growth of speech. 
We at once see how hopeless must be the quest of a primitive 
mother-tongue, which never existed, the faculty starting from a 


germ and developing itself in different regions independently. 
Thus also is exposed the fallacious assumption of speech being 
"created" or consciously "invented" by primitive man himself, 
and then passed on from one tribe to another, as when even 
M. Letourneau writes : " We may perhaps infer that these races 
have not created their own languages, and that during the very 
long prehistoric period foreign initiators brought to them idioms 
which had taken root and grown elsewhere" (Sociology , p. 581). 
It must now be obvious that no speechless people could be 
taught to talk a ready-made language unless they possessed "the 
necessary physical organs, in which case they would not need 
to be taught, being already in possession of a language of their 
own. The organs and the faculty must have been developed 
simultaneously by repeated tentative and unconscious efforts. 

In the same way many other abstruse questions connected 
with the natural history of man his physical and menial evolu- 
tion, his antiquity, his specific unity and varietal diversity seem 
to pass easily from the field of abstract speculation to that of 
solid fact, when approached from the evolutionary standpoint. 
From any other standpoint they remain, as before, either hopeless 
riddles or the sport of theological theorists and metaphysical 
dreamers. What, therefore, might here appear at first sight too 
assertive and over-confident, may on reflection be found the 
simple, often the inevitable, outcome of inductive reasoning. 
When, for instance, Prof. Prestwich speaks of " 20,000 or 30,000 
years " as the extreme length of man's days on enrth, can it 
be rash to unhesitatingly reject such a narrow estimate in the 
face of the daily accumulating evidences of his vast antiquity 
brought to light by such competent explorers as Mr Worthington 
Smith, Mr W. J. Lewis Abbott, Mr H. Stopes and Mr Harrison 
of Ightham in the present Thames basin, Dr Collignon and 
Dr Couillault in North Africa, Prof. Flinders Petrie in the 
Nile Valley, Prof. Sergi in Italy, Herr Maschka in Bohemia, 
Dr Noetling in Indo-China, Sig. Lovisato in Fuegia, Mr W. H. 
Hudson in Patagonia, Dr C C. Abbott and*Mr E. D. Cope in the 
United States, and others elsewhere? While proofs are being 
collected of pliocene, and even "early pliocene'' man, and while 
Dr Dubois' Pithecanthropus credits supplies a distinct missing link 


between the anthropoids and the Neanderthal race, does not the 
rashness lie rather with those who would limit the age of eolithic 
man to late or even post-pleistocene times ? When it is remem- 
bered that fully 8,000 years ago the Egyptian language was not 
only developed but already entirely severed from its original con- 
nection with the Semitic group, it becomes obvious that merely 
to account for the highly specialised Hamito-Semitic division 
a much longer period will be needed than is conceded by 
Prof. Prestwich to the human family itself. Hitherto " Tertiary 
man* not proven " has rightly been the watchword of the English 
conservative school of ethnologists. May it not be asked whether 
the negative particle should not now be struck out of this formula, 
seeing that almost without exception their continental fellow- 
workers have with Virchow surrendered the point, and that strong 
evidence of pliocene man has been brought forward by Sergi 
in Italy and Noetling in Burma, if not also by Stopes and 
Harrison in Kent ? As these lines are being penned Mr Stopes 
reports from Swanscombe near Gravesend and from Ash a few 
miles farther south, numerous finds belonging to all ages, " from 
the British back to that very remote period when the gravels 
were being deposited on the high plateau of Kent in pliocene 
times" (Athen&um, Sept. 7, 1895, p. 325). 

Clearness has been consulted by a twofold division of the 
subject-matter, the first dealing with those fundamental problems 
which affect the human family taken as a whole, the second 
discussing the more general questions which concern the Homi- 
nidae, that is, the several main branches of mankind. In the 
first division are introduced some topics, such as the physical 
evolution of man, his points of contact with the other groups 
of primates, and the physical criteria of race, which might seem 
to belong more properly to the field of special anthropology. 
But in all closely allied branches of knowledge such encroach- 
ments necessarily occur, as, for instance, when geography tres- 
passes on geology, and geology on astronomy. In the present 
instance the " trespass " will perhaps be all the more welcome 
because no comprehensive work on special anthropology has yet 
appeared in the English language, so that the student is still 
mainly dependent on Topinard's masterly treatise. In any case 


the introduction of certain anthropological matters was inevitable, 
the mental qualities, of which special anthropology takes no 
account, being largely determined by the physical constitution, 
just as the mind itself has its seat in a physical organ. "Although 
Mind can never be identified with Matter, nor the acts and 
states of the mind reduced to acts and states of the brain, yet 
as the latter are the physical antecedents of the former, the 
study of the one class of phenomena is calculated to give light 
and guidance in the study of the other " (Dean Byrne, General 
Principles of the Structure of Language, n. p. 380). And more 
pointedly elsewhere: "Though thought be not regarded as a 
function of the brain, yet it is the function of the brain to 
minister to the acts of thought, so that cerebral action is the 
condition of mental action. Between these two actions there 
must be an exact correspondence ; so that both must be studied 
if we would understand either" (p. 379). 

In the second part those general questions alone are treated 
which concern the primary human groups. Here the main object 
has been to solve some of the more fundamental problems con- 
nected with these groups, and thus clear the ground for a complete 
classification of the Hominidae. But no attempt is made at such 
a classification which would require a work to itself, and which 
may form the subject of a future volume of the present series. 

Meantime, a hope may be expressed that this summary of 
ethnological data will be found helpful to the student, by en- 
abling him to group and coordinate his facts, to understand their 
mutual bearings, and to fit them into their proper place in the 
natural history of the human family. But, above all, it should 
teach him to reason correctly, and draw the right inferences from 
established premisses, at whatever cost to biassed or precon- 
ceived theories on the fundamental ethnical problems. Thus 
alone can a hope be entertained of some law and order being 
introduced into the present chaotic state of the public mind on 
all matters connected with "man's place in nature. " In his 
monograph on Sculptured Anthropoid Ape Heads from Oregon 
(New York, 1891), Mr James Terry draws a deplorable picture of 
American anthropological literature, "already so filled with op- 
posing theories that it appals the student who undertakes to 


unravel the contradistinctions [contradictions?] of its many writers." 
But the New World can pretend to no monopoly of such bewil- 
dering conflicts of opinion. That Mr Terry's picture admits of 
wider application is made only too evident by a glance at the 
wild theories of emotional ethnology still persistent amongst our- 
selves, theories supported by the reckless comparisons and con- 
clusions even of capable writers, who, in the absence of accepted 
first principles, give bridle to their imagination, and replace sober 
reasoning by extravagant speculation. Thus whole populations 
Japanese, Malays, Egyptians are, so to say, transferred bodily 
from the Eastern to the Western Hemisphere, in order to account 
for shadowy resemblances between the cultures of the Old and 
New Worlds. And, as if to show the absurdity of this line of reason- 
ing, Dr A. le Plongeon now proposes to reverse the process and 
make "Mayax" [Yucatan] the "hub of the Universe." Develop- 
ing the ideas tentatively advanced in his Sacred Mysteries among 
the Mayas and Quiches 11,500 years ago (New York, 1886), this 
antiquary boldly places the cradleland of mankind itself in Central 
America, where he discovers the tomb and the very dust of Abel 
slain by Cain, and even "the very weapon employed in the crime." 
Here, we are told, is still spoken the stock language which affords 
a key to the interpretation of ancient Egyptian, Sanskrit and 
Hellenic formulas, while the Greek alphabet itself is shown to be 
merely an epic poem on the Cain-Abel legend, composed in the 
same primitive Maya tongue. Even the letters of late introduction, 
such as epsilon, omikron, omega, bearing pure Greek names, do not 
escape this philological crucible, omikron, for instance, being re- 
solved into the Maya elements om = whirlpool, /#=wind, fa = place, 
and on Around, meaning "whirlwinds blow round." Ample details 
of these "startling revelations," divulged in all seriousness, are 
communicated through Mr O'Sullivan, H.B.M. Vice-Consul at 
Pemba, to the Review of Reviews for September, 1895, and thus 
acquire a sort of official stamp. 

Another case in point is the rivalry still maintained between 
many prominent exponents of the anthropological and philo- 
logical sciences, whose antagonism has flooded ethnological 
literature with barren controversy, and retarded the progress of 
these sister sciences by confused methods of ratiocination. It 

PREFACE, xiii 

contended, on the one hand, that the races of men spring 
>m several geographical centres independently, because their 
iguages are fundamentally distinct ; it is retorted on the other 
,nd that language and race have nothing in common, or at least 
e in no way correlated. But when the nature of the evidence 
examined in the light of the first principles which the present 
)rk aims at establishing, it is seen that neither of such extreme 
?ws can be right, while a way may nevertheless be found to 
-oncile the rival claims of the anthropologist and the philologist 
!haps. VIL and IX.). From this example we see how trq it 

that an essential condition for the successful prosecution of 
hnological studies is the power of reasoning aright on the 
:fs admitted and appealed to by both sides. 

But a more formidable rivalry, and one destined probably to 
>t longer, is that which persists between dogmatism and the 
ological sciences. In his presidential address at the meeting of 
e British Association, Ipswich, 1895, Sir Douglas Galton referred 

the services rendered to the advancement of knowledge by the 
te Professor Huxley, whose action had helped to sweep away the 
>structions of dogmatic authority which in the early days of the 
ssociation had fettered progress, especially in anthropological 
.idies, and whose energy and wealth of argument had largely aided 

winning the battle of evolution and securing the right to discuss 
icstions of religion and science without fear or favour. The 
>mage paid to the memory of the great captains on the scientific 
ie, the greatest of whom found a resting-place in the British 
alhalla, warrants the belief that their opponents are now willing 

give f their arguments at least a fair hearing. When it is further 
en, as the late Professor J. D. Dana clearly saw, that there is 
>thing in the doctrine of evolution rightly understood " to impair 

disturb religious faith" (Letter to the Rev. J. G. Hall, Cleve- 
nd, Ohio, March 3, 1889), we shall have arrived at a measurable 
stance of the time when that doctrine will take its place by the 
ie of the Copernican and Newtonian teachings, as an elementary 
ith at the foundation of a rational conception of man and the 
liverse. Then a way will also be found, as already here sug- 
:sted (p. 30), to reconcile the views of Science and Religion on 
e origin and evolution of the human species. But it would be 


idle to pretend that there can be any compromise on the part of 
Science. Hence such a reconciliation must necessarily involve 
some concessions by the dogmatists, such, for instance, as enabled 
them to ultimately accept the Copernican view of the solar system, 
despite the geocentric theory prematurely raised to a dogma on 
the strength of Biblical texts. Such developments within the 
sanctuary are inevitable if the religions are to retain the respect of 
their more thoughtful adherents, and British orthodoxy itself is 
warned by the present head of the Church of England not to 
forget " that every age does and ought to shed new light on truth. 
To refuse to admit such light and its inherent warmth is to forfeit 
the power of seeing things as they are and to lose the vigour of 
growth. It is, in fact, to limit ourselves finally to a conventional 
use of hard formulas " (Pastoral Letter, August 30, 1895). 

In a work of this nature, dealing with a multiplicity of subjects, 
on all of which nobody can be supposed to have personal know- 
ledge, it is not to be expected that the views advocated, or even 
the mere statements of facts, will be always accepted on the ipse 
dixit of the writer. Hence the necessity of constant reference to 
received authorities, which may possibly here and there encumber 
the text, but which will not on that account be objected to by the 
serious student. Quotations, however, especially from foreign 
sources, are in most cases transferred to the footnotes, where the 
reader will find nearly ail important statements supported by 
proof or authority of some kind. At the same time full responsi- 
bility is accepted for all theories or conclusions which are here 
advanced for the first time, or which at least are not known by 'the 
author to have been put forward by any previous writer. Such are 
in Part I. the evolution of neolithic megalithic architecture (Chap. 
VI.); the relation of stock languages to stock races (Chap. VII.); 
the evolution of the various morphological orders of speech, and 
the general relations of race and language (Chap. IX.); in Part II. 
the order of evolution of the primary groups, and their centre of 
evolution and dispersion (Chap. X.) ; the treatment of the linguistic 
problem in Oceanica, and of the racial problem in Australia and 
Tasmania (Chap. XL); the Finno-Tartar, Chukchi and Malay 
racial problems, and the Malayo-Polynesian linguistic relations 
(Chap. XII.) ; the peopling of America in the Stone Ages, the 


independent local evolution of American cultures, and the treat- 
ment of the Eskimo question (Chap. XIII.) ; the general treatment 
of Homo Caucasicus, the Ibero-Berber question, the Aryan cradle- 
land and the Aryan race problem (Chap. XIV.). 

It remains gratefully to acknowledge the loan of photographs 
and other illustrations, elsewhere specified, from Messrs Flower 
and Lydekker, Sir John Evans, Dr D. J. Cunningham, the pub- 
lishers of Nature^ Mr Edward Stanford, Messrs Longmans, Mr 
J. J. Lister, Dr H. O. Forbes, Mr W. T. Stead, and the Royal 
Geographical Society. Thanks are also due to Messrs Cassell for 
their kind permission to use some of the ethnological material 
contributed by the writer to their Storehouse of Information^ and 
to the Editor of this series, whose careful revision was not confined 
to typographical matters. 

A. H. K. 


October ; 1 895. 







Definitions- Anthropology General and Special Ethnology Ethnography 
Scope of Ethnology General Nomenclature Definite Terms : Rare ; 
Clan; Tribe; Family ; Totem; Branch; Stock; Type Indefinite Terms : 
Division; Section; Group; Horde; Nation; People Example i 15 



Man's Place in the Animal Kingdom The Primates- Old Divisions: Quad- 
rumana and Biruana New Divisions: Lemuroidea and Anthropoidea 
The five families of the Anthropoidea Their range in time and space 
Diagram of the Anthropoid families Relations of the family Hominidx* to 
the family Simiida? Comparative Table of the Simiidu: and ITominidse: 
Gibbon; Orang; Gorilla; Chimpanzee; Dryopithecus ; Horninidce 
Points of resemblance to and difference from the Simiidzc Origin of 
Man by Creation or Evolution Creation Theory inadequate Evolution 
Theory adequate and Supernatural x r iews reconciled Difficulties 
of the progressive evolutionist theory Views of de Qualrefuges, dc 
Mortillet and Sergi The Castenedolo Man Sergi's Tertiary Hominirige 
Quaternary Man Cannstadt Man rejected- Neanderthal affirmed The 
Quaternary Hominidac -Kollmann's Dauertypus Persistence of primitive 
types Views of French, English and American Anthropologists Diffi- 
culties of the Dauertypus theory Analogy of thu Equidue Their evolution 

xviii CONTENTS. 

Sergi's Tertiary Hominidse rejected Persistence of, and Reversals to, 
primitive types reconciled with evolutionary teachings Comparative 
Diagrams of Pleistocene Hominidse and Equidae Broad stages of physical 
evolution from a postulated Anthropoid Miocene precursor . 16 39 



Wuman incomparably greater than animal intelligence Growth of mind 
apparently out of proportion to that of its seat, the brain Evolution of 
organ and function correlated Cranial to be distinguished from mental 
capacity Comparative cranial studies often contradictory Chief physical 
determinants of mental power not so much the volume of the brain as 
its convolutions and the cellular structure of the grey cortex These 
elements capable of indefinite expansion till arrested by the closing of the 
cranial sutures Different degrees of intelligence in different races accounted 
for Such differences independent of the general bodily structure Hence 
physique and mental power not necessarily correlated and not always 
developed pari passu But mind and cerebral structure always corre- 
spond Hence comparative study of brain texture, as by Broca and 
Miklukho-Maclay, yield best results Brain and its function, thought, 
capable of indefinite future expansion Differ in degree only, not in kind, 
from those of the lower orders Time alone needed to bridge the gap 




The Geological Sequence in its bearing on the Antiquity of Man Table of the 
Geological Sequence: Primary; Secondary; Tertiary; Quaternary; Pre- 
historic; Historic The Glacial Problem Reactionary Views Croll's 
Periodicity Theory Objections and Limitations of Time by Prestwich A 
reductio ad absurdum Arguments based on influence of Gulf Stream and 
Absence of Glaciation in earlier geological epochs estimated Croll's Theory 
reaffirmed A long period of time needed to meet all the conditions : Re- 
distribution of Land and Water; Intermingling of Arctic and Tropica. 
faunas; Scouring out of great river valleys; Man long associated with 
extinct animals ; Britain twice submerged since its occupation by man ; 
Little trace of primitive man in the last post-glacial deposits of the North 
Two Ice-ages and long Inter-glacial period essential factors Difficulties 


of the Intermingled Arctic and Tropical Faunas Lyell and Boyd-Dawkins* 
"Seasonal- Migration" Theory discussed Long association of reindeer 
and hyaena explained Great age of the flints found in the high-level 
drift, boulder-clay, plateaux and riverside terraces Pre-, Inter- and Post- 
Glacial Man The problem restated General Conclusion Pliocene 
Hominidse rejected Specialised Inter-glacial Hominidae reaffirmed 
Their probable age Post-glacial Man a nondescript . . 50 70 



Palaeolithic Man spread over the whole world But in many places early and 
later Cultures run in parallel lines, not in time sequence Hence the Time 
relations often obscured, objects of human industry not being everywhere 
tests of age, but only of grades of culture Even these grades not always 
clearly distinguished Palaeolithic art not stationary but progressive, and 
in some respects outstripping that of neolithic times Materials available 
for the study of primitive Man: implements, monuments and human 
remains Unreasonable objections to implements (palaeoliths) as evidence 
of antiquity Value of implements determined by their provenance and 
associations in geological formations or in caves Stalagmite beds not 
necessarily a test of age Kitchen-middens of all ages, some very old, 
some recent and of rapid growth, hence to be judged on their merits 
Human remains reserved for special treatment Quaternary Man in 
Britain: Evidence of Halfield Beds; Kent's Cavern; Brixham Cave; 
Cresswell and Victoria Caves ; Lotherdale and Pont Newydd Caves ; Vale 
of Clwydd Caves ; Thames river-drift ; High-level gravels ; Chalk plateau, 
Kent ; Eoliths from Canterbury gravels, Stoke Newington, &c. Quater- 
nary Man in France : Somme Valley river-drifts, St Acheul Grades of 
Palaeolithic Culture De Mortillet's Four Epochs: Chellian Age, typical 
implements ; Moustierian Age, typical implements ; Solutrian Age, typical 
implements ; Madelenian Age, typical implements The Dordognc School 
of Art Placard Cave : Superimposed Culture eras Evidences of Palaeo- 
lithic Man in France and Italy Quaternary Man in Africa (Egypt, 
Algeria, the Cape); in Asia (Syria, Palestine, Asia Minor, Caucasus, Mon- 
golia, India, Japan) ; in Australia and New Zealand ; in America (Tierra 
del Fuego, Patagonia, Argentina and Brazil, Mexico, United States and 
Canada) ; evidence from the Trenton gravels ; Mississippi Basin and other 
localities ; Views of Chamberlin, Holmes, Mason Ind other conservatives 
on the value of this evidence ; the Calaveras Skull General Diffusion of 
Primitive Man throughout North America The Mound-builders not 
quaternary; their Culture neolithic, prehistoric and historic 71 107 




Marked difference between the Old and New Stone Ages Comparative Table 
of Palaeo- and Neolithic Cultures A Break of Continuity in some regions, 
notably in Britain But not everywhere No universal hiatus possible 
Continuous evolution in the south and south-east Probable duration of 
neolithic times The late palaeolithic era of the West synchronous with 
the early neolithic era of the South-east Great duration of neolithic 
times argued on general considerations The Danish peat-bogs a time 
gauge The Danish kitchen-middens Origin and growth of aquatic 
stations The Swiss Lake- Dwellings The Irish and Scotch crannogs 
Neolithic structures Reducible to two types: The polylith or cell, and 
monolith or block, originating in Burial and Ancestry worship Polylithic 
and monolithic nomenclature Evolution of the Cromlech or Dolmen 
through the Barrow from the Cell Popularly associated with druidical 
rites The Sessi and Stnzzone of Malta and Corsica The Nuraghi of 
Sardinia The Talayots of the Balearic Islands The Russian Kurgans 
Silbury Hill The Cell becomes a Family Vault with later develop- 
ments The Menhir, its origin and wide diffusion Its development in 
linear and circular direction The Alignments and Cycloliths (Stone 
Circles) Their origin and purpose explained Erdeven ; Stonehenge ; 
New Grange ; Menec, Carnac district The Irish Round Towers 
Geographical Distribution of the Megaliths Chief Centres: Bahrein 
Islands; Moab; Mauritania; Gaul, Britain, Scandinavia Bearing on the 
question of early migrations Europe re-settled in Neolithic times from 
two quarters Routes indicated by the presence or absence of Mega- 
lithic Structures These wrongly accredited to the Kelts who followed 
the non-megalithic route Astronomic and religious ideas attributed to 
the megalith-builders Prehistoric monuments in the New World General 
Survey Tiahuanaco, culminating glory of American Megalithic archi- 
tecture Tiahuanaco Culture an independent local development 108 140 



Specific or Varietal unity decided by extent of divergence between past and 
present races Species and Variety The Physiological test : inter-racial 
fertility The Canidte, Equidae and Hominidae The Palaeolithic races 


Their remains : Trinil : Homo Neanderthalensis ; LaNaulette; LaDenise; 
Spy; Kent; Podbaba; Predmost; Marcilly; Mentone; Olmo; Eguis- 
heim; Laugerie; Palaeolithic races exclusively long-headed Neolithic 
races at first also long-headed, then mixed, and later exclusively round- 
headed in some places But all intermingled Fertile miscegenation 
established for prehistoric times In the historic period mixture the rule, 
racial purity the exception The Mestizos of Latin America The Paulistas, 
Franco-Canadians, and Dario- Eskimo The United States Indians and 
half-breeds Eugenesis established for the New World, and for Africa : 
The Griquas, Abyssinians, Sudanese, and West African Negroes Mixed 
races in Asia, Malaysia, and Polynesia The Pitcairn Islanders The physio- 
logical test conclusive against the PolygenistsThe anatomical test ^he 
Polygenist linguistic argument: Independent stock races inferred from 
independent stock languages Fallacy of this argument Specific Unity 
unaffected by the existence of Stock Languages which are to be other- 
wise explained The Monogenist view established and confirmed by the 
universal diffusion of articulate speech Psychic argument The question 
summed up by Blumenbach .... . 141 161 



Difficulties of defining, and determining the number of, the primary human 
varieties Schemes of the first systematists : Bernier ; Linne ; Blumenbach ; 
Cuvicr; Virey ; Desmoulins ; Bory de Saint- Vincent ; Morton; GHddon 
and Agassiz; Latham; Carus; Peschel The Philologists The Ethno- 
logists : Buffon ; Prichard The Anatomists : Geoffroy Saint-Hilaire ; 
Retzius ; Broca ; Virchow ; Mantegazza ; Barnard Davis ; Kolleston ; 
Flower; Cope Recent Schemes : Haeckel's; de Quatrefages's ; Huxley's; 
Broca's; Fr. M tiller's ; Deniker's; Flower and Lydekker's General 
remarks on these Groupings Elements of Classification: Physical and 
Mental Characters Physical tests of Race : Colour of the SkinColour 
and Texture of the Hair The Beard ; Hirsuteness Shape of the Skull 
Cephalic Indices Tables of Dolicho-, Mesati- and Brachycephali 
Gnathism Facial Index Table of Sub-nasal PrognathismThe Denti- 
tion The Nose: Nasal Index Colour and Shape of the Eye The 
General Expression Stature : Tables of Heights-yOther Physical Factors 





Cranial Capacity Size of brain and Mental Capacity correlated in the animal 
series and partly in man Comparative Tables of Cranial Capacity 
Language the chief mental criterion Relation of speech to Anthropology 
Phonesis a physical function which cannot be neglected by the anthro- 
pologist Value of language to the ethnologist Evolution of speech 
from the inorganic to the organic state The faculty originated most pro- 
bably in a single centre Reply to the linguistic polygenists Speech of 
relatively recent growth Hence at first unstable and subject to great 
fluctuations Hence also linguistic divergence more rapid than physical 
types, forming species and genera which cannot mix Hence no mixed 
languages Consequent value of speech as a racial test Linguistic more 
easily distinguished than physical groups Table of mixed peoples speak- 
ing unmixed languages Table of peoples whose speech has shifted with- 
out mixing Table of peoples whose physical type has changed, their 
speech persisting Hence speech and race not convertible terms But speech 
often a ^reat aid in determining ethnical elements The morphological 
orders of speech Old views of linguistic growth The "Root" theory 
Monosyllabism not the first but the last stage of growth The sentence 
the starting-point The monosyllabic languages originally polysyllabic 
Chinese the result of phonetic decay The Aryan root theory exploded 
Root and Atom; Sentence and Molecule Agglutination Its nature and 
test The morphological orders not fixed species but transitional phases 
of growth Inflection reverts to Agglutination Agglutination passes into 
Inflection and Polysynthesis Polysynthesis not a primitive but a late 
condition of speech Differs in kind from Agglutination Nature of In- 
flection Diagram of linguistic evolution Development of speech not 
linear but in parallel lines Synthesis and Polysynthesis tend towards 
monosyllabic analysis Change from pre- to post-position in the Aryan 
group Change the Universal Law of all living speech Social state : Fish- 
ing, Hunting, Agriculture, no test of race Social Usages poor criteria 
Religion Origin and development of nature and ancestry worship- 
Anthropomorphism due to the common psychic character of man Hence 
common religious ideas no proof of common origin or of contact Like 
usages no evidence of common descent 190319 





Four Primary Groups Homo ^Ethiopicus, Mongolicus, American us, Caucasi- 
cus Family Tree of the Hominidae The primary groups derived, not 
one from the other, but independently from a common precursor Their 
differences determined by their different environments Position of the 
several groups The Negro The Mongol and American The Caucasian 
Remarks on this Terminology Comparative Table of the physical and 
mental characters of the four primary groups Centre of Evolution Dis- 
tribution of land and water in Secondary and Tertiary Times The Indo- 
African Continent The Austral Continent The Eurafrican Continent- 
The Euramerican Continent America accessible from Europe and from 
Asia Theory of de Quatrefages on the migrations of primitive man His 
linguistic argument Views of Dallas and Urinton Evolution "with a 
jump" The Missing Link Probable centre of Evolution and Dispersion 
the Indo-African and Austral regions, true Home of the Lemurs and of 
the Anthropoids Characters of the pliocene precursor and of the pleisto- 
cene sub-groups persistent in the Afro-Austral regions Pliocene and 
pleistocene migrations from the primeval home Order of Development 
of the primary groups in their several centres of evolution Monogenist 
and Polygenist views reconciled Flower and Lydckker on the spread of 
the Hominidae over the globe 221 241 



Two divisions: African and Oceanic Negro Family Tree The Negritoes: 
Two divisions Early migrations The African Negritoes The Akkas 
and BatwaThe Bushmen and Hottentots Past and present Hottentot- 


Bushman domains The Oceanic Negritoes The Black element in India 
The Oceanic Negrito groups : Andamanese ; Sakais of the Malay 
Peninsula; Aetas of the Philippines; Karons of New Guinea; Kalangs 
of Java The Negro divisions compared The African Negro unprogressive 
without miscegenation Testimony of H. H, Johnston, Manetta, Ruffin 
and Sir Spencer St John Historic evidence Low state of Negro culture 
Two main sub-divisions: Sudanese and Bantu The Sudanese Negroes 
Mixed Sudanese groups The Fulahs The Negroid Bantus The Zulu- 
Kafirs and Wa-Huma The Bantu linguistic family General intermingling 
of the Sudanese and Bantu populations Hence classification impossible 
except on a linguistic basis Tables of the Sudanese and Bantu groups 
The Oceanic Negro domain An area of great ethnical confusion Two 
main sub-divisions: Insular Negroes and Negroid Australians Nomen- 
clature: Melanesians; Papuans The Papuan domain, past and present 
The Papuan type The linguistic problem Wide diffusion of Malayo- 
Polynesian speech not due to Malay or Polynesian Migrations Still less 
to Melanesian Migrations The true explanation; the Caucasic factor- 
The Australian sub-division Not homogeneous Constituent elements of 
the Negroid Australians and of the Tasmanians Tasmanian culture 
eolithic 242 -294 



home of the Mongol race easily accessible to the pliocene precursor 
Transition from the generalised human type to the Mongol variety 
Chief Mongol physical characters Diffusion of the Mongol race Early 
Mongolo-Caucasic interminglings Hence aberrant Mongolic groups 
Mongol Family Tree Chief Mongol sub-divisions Their domain The 
Akkads Early linguistic relations The Mongolo-Tatar sub-division 
Nomenclature : Mongol ; Tatar ; Turki Divergent Finno-Turki types 
The Samoyedes The Lapps The Baltic Finns ; Karelians ; Tavastians 
White elements in the Mongolo-Tatar domain Avars Magyars 
Bulgars Osmanli affinities Koreo-Japanese group The Koreans The 
Japanese: Physical qualities; Mental qualities The "Hyperboreans" 
The Chukchi problem The Tibeto-Indo-Chinese sub-division General 
physical uniformity Tibeto-Chinese linguistic relations Function of 
Tone in the Isolating Languages Tibetan linguistic affinities Indo- 
Oceanic linguistic relations The Indonesians The Malay problem 
Malay physical type Malagasy affinities Malayo- Polynesian linguistic 
relations Ethnical relations in the Philippine Islands. . 295 333 




America peopled from the Eastern Hemisphere during the Stone Ages The 
bronze age of Chimu (Peru) no proof of later intercourse between the 
Old and New Worlds Hence the American aborigines are the direct 
descendants of palaeolithic and neolithic man and their later culture is 
consequently an independent local development But Homo Amerioanus 
is not autochthonous, but a specialised form of a Mongol prototype 
General Uniformity of the American physical type Texture of the hair ; 
colour of the skin "White" and " Black" aborigines no proof of 
early migrations from Europe or Melanesia Arguments of De Quatrefages 
discussed The Japanese myth exposed The "stranded junk " argument 
Culture of the early Stone Age identical in both hemispheres But 
after that age the arts and industries show continuous divergence in 
America Argument based by Retzius on the two types of American 
crania Contrasts between the .present Mongol and American physical 
types Mental Capacity of the American aborigines superior to the 
Negro, on the whole inferior to the Mongol But the Cranial Capacity 
inferior both to Mongol and Negro Striking uniformity of the mental 
characters of the aborigines in North America in South America 
Uniform character of American speech in its general morphology 
Fundamentally distinct from the structure of the languages of the Old 
World Surprising number of American stock languages despite their 
common polysynthetic type Classification of the aborigines must always 
be mainly based on language Family Tree of Homo Americanus 
America probably peopled by two routes From Europe by palaeolithic, 
from Asia by neolithic man Present distribution of the two types 
The Eskimo question Its solution Prof. Mason's theory of the peopling 
of America from Indo-Malaysia Negative Objections to this theory 
Positive Objections True explanation of the coincidences between 
certain usages and mental aspects of the inhabitants of the Old and 
New Worlds Due not to contact or borrowings, but to their common 
psychic constitution Results of the discovery and re-settlement of 
America on the aborigines in Latin America In Anglo-Saxon America 
The Anglo-American type due, not to miscegenation, but to con- 
vergence 334373 




North Africa probable cradle of the Caucasic racewhich spread thence east 
to Asia and north to Europe The Cro-Magnon and other early European 
races affiliated to the fair Berbers of Mauritania West Europe occupied 
by several varieties of Homo Caucasians in the Stone Ages Who were of 
Son- Aryan speech like the still surviving Basques The Ibero- Berber 
problem Basques and Picts Family Tree of Homo Caucasicus 
Xanthochroi and Melanochroi Blacks of Caucasic Type Physical 
Characters of Homo Caucasicus White, Brown and Dark Hamites 
The Tamahu Hamites of the Egyptian records The "New Race" in 
the Nile Valley The Eastern Hamites: Afars; Bejas; Gallas and 
Somals ; Masai and Wa-Huma Ethnical relations in Abyssinia : Him- 
yarites ; Agaos ; The present Abyssinian populations Relations of the 
Hamites to the Semites The Semitic Domain The Semitic Groups 
Semitic physical and mental characters The Semitic Languages The 
Aryan-speaking Peoples Aryan a linguistic not a racial expression- 
True character of the Aryan migrations Illustrated by the Teutonic in- 
vasion of Britain; and by the Hindu invasion of India The Aryan 
Cradleland Primitive Aryan Culture Schrader's hypothesis Conflicting 
views regarding the Aryan Cradleland reconciled The Eurasian Steppe 
true home of the primitive Aryan Groups The primitive Aryan type 
difficult to determine But probably xanthochroid The Aryan problem 
summed up- Recent expansion of the Aryan-speaking Peoples The 
"Greater Britain" The Aryan linguistic family Table of the Aryan 
linguistic groups Disintegration of primitive Aryan speech The Teu- 
tonic phonetic System Ethnical and linguistic relations in the Caucasus 
Main Divisions of the peoples and languages of Caucasia Ethnical 
and linguistic relations of the Dravidas Sporadic Caucasic Groups: 
Todas ; Ainus 374420 

Ai'PKNPix 4214-26 

INDEX . 427442 



Diagram of the Anthropoid suborder of Primates 19 

Skull of Orang, from Flower and Lydekker's Mammals Living and 

Extinct ............ 21 

Orang- Utan, from Guillemard's Cruise of the Marchesa . . . .21 

Chimpanzee, from Lydekker's Royal Natural History .... 23 

Diagram of the Simiidye and Man, reduced from Huxley's Marts Place 

in Nature 27 

The Neanderthal Skull, from a photograph 33 

Diagram showing the Evolution of the Kquidse, from Flower and 

Lydekker, op. cit . .36 

Comparative Diagrams of the Pleistocene Hominidoe and Equidae . . 38 
Remains of Palaeolithic Man from Kent's Cavern, from Sir John Evans' 

Ancient Stone Implements of Great Britain 79 

River Drift Palseolith from Santon Downham, ibidem .... 80 

River Drift Palaeolith from Red hill, ibidem 85 

Palaeolithic Engravings from Duruthy and La Madeleine Caves . . 88 

The Placard Cave, with Section of Floor 90 

Palaeoliths from ihe District of Colombia, U.S., from Wilson's Prehistoric 

Anthropology .......... 102 

Neoliths from various localities in the United States, ibidem . . ,109 
Neolithic Celt from Bridlington, from Sir J. Evans* Ancient Stone Imple- 
ments 112 

Neolithic Arrow-head from the Yorkshire Wolds, ibidem . . . 113 

Neolithic Stemmed Arrow-head from the Yorkshire Wolds, ibidem. . 114 

Neolithic Javelin or Arrow-head from Iwerne Minster, Dorset, ibidem . 116 

Neolithic perforated Axe from Hunmanby, Yorks, ibidem . . . n8 
Trevethy Stones, from a photograph . . . * . , . .125 

Ground-plan of Palo-de-Vinha Dolmen, Portugal 126 

Dolmen-Tumulus of Kercado, Brittany 128 

Dol Menhir, from a photograph , 1 29 



The Carnac Alignments, from a photograph 132 

Skull of Pithecanthropus crcctus, from Dr E. Dubois' Monograph, 

Batavia, 1894 144 

Comparative Diagram of Irish, Spy, Neanderthal, Pithecanthropus and 
Gorilla Crania, from Dr D. J. Cunningham's Paper on Pithecan- 
thropus, Nature^ Feb. 28, 1895 145 

The Spy Cranium No. i, from Ph. Salmon's Races Humaines Prt- 

historiques 146 

Diagram of J, Deniker's Scheme of Classification of Races, BtiL de la 

Soc. d 1 * Anthrofologie, June, 1889 169 

Diagrams showing the various forms of the human hair in transverse 

section 176 

Orthognathous Skull of Kalmuk, after von Bauer . . . . . 1 83 

Prognathous Skull of Negro 183 


Family Tree of the Hommidre 224 

Family Tree of Homo /Ethiopicus ....... 244 

Akka of Mangbattuland (African Negrito Type) ..... 247 

Sakai of Malay Peninsula (Oceanic Negrito Type), from a photograph by 

Miklukho-Maclay 258 

Samang of Malay Peninsula (Oceanic Negrito Type), from a photograph 

by Miklukho-Maclay 258 

Aeta Woman of Luzon (Oceanic Negrito TyP e )> from a photograph in 

A. B. Meyer's Album von Philippintn-Tpycn, Dresden, 1885 . 260, 261 
Ardi, a Kalang of Java (Oceanic Negrito Type), from a photograph by 

H. van Musschenbroek 262 

A Zulu Girl of Natal (Bantu Type), from a photograph . . . .271 
Susu Negro, Senegambia, from a photograph by Prince Roland Bonaparte 276 
Aduma Negro, Ogoway Basin, from a photograph by Prince Roland 

Bonaparte 278 

Australian (normal Type), from A, H. Keane's Types of the Races of 

Mankind, Longman's New Atlas, 1889 280 

Native of Duke of York Island (Melanesian Type), from a photograph by 

O. Finsch (Rtisc in der Stft/see, Berlin, 1 884) 282 

Native of New Britain (Melanesian Type), from a photograph by O. 

Finsch, ibidem 282 

Native of .Nifelole Island (Melanesian Type), from a photograph by the 

Rev, W. G. Lawes 282 

Native of S. E. New Guinea (Papuan Type), from a photograph by 

H. 0. Forbes 283 

Native of Dutch New Guinea (Papuan Type), from Guillemard's Austral- 
asia (Stanford Series) 286 



Australian of Queensland (primitive Type), from a photograph by J. J 

Lister ........ .... 393 

Native of Tasmania (normal Type), from a sketch by Lieut. F. G. S. de 

Wesselow, R.N ........... 393 

Manchu of Kulja (full face) \ from a photographic album of Central 
- , ,, (profile) I Asian Types taken at Tashkend in 

Kalmuk Woman (West Mongol | 1876; R. Geograph. Society's Col- 

Type) * lection ...... 198 

Family Tree of Homo Mongolicus ........ 300 

Akkad of Babylonia (Mongol Type?), restored by Theo. Pinches (Types 

of the Early Inhabitants of Mesopotamia in Joum. Anthrop. Imt* 

1891,?. 91) ... ....... 301 

Ostyak Woman (North Mongol Type), from S. Sommier's Siri<mi> Ostiacchi 

e Samoicdi) Florence 1887, p. 88 ....... 306 

Kara-Kirghiz of Semirechinsk (full face) ' 

M n ,, (profile) 

Kara-Kirghiz Woman of Semirechinsk , , _ , , 
Kirghiz of Tashkend (Turki Type) from lhe Ta ' hke d Album of 

Uzbeg of Zerafshan District (Mixed 

Photographs (see above) 310, 3 

Turko- Iranian Type) 
Solon of Kulja (Manchu Type) 

Japanese Woman, from a Japanese photograph 315 

Japanese Jinricksha runner, from a Japanese photograph . . '3*5 
Siamese (Indo-Chinese Type) | from A. H. Keane's Types of the. Races of 
Chukchi of N.E. Siberia j Mankind (see above) . . . . 318 
Annamese of Saigon (Indo-Chinese Type), from a photograph by Prince 

Roland Bonaparte . . . . . . . . . . 370 

Burmese Lady (Indo-Chinese Type), from a photograph . . . 320 
Chinese Woman of Kulja (full face) "j from the Tashkend Album (see 

,, ,, ,, (profile) J above) 311 

Sundanese of West Java (Malay Type), from a photograph by Prince 

Roland Bonaparte . . . . 327 

Native of Tonga Is. (Eastern Indonesian Type), from Guillemard's 

Australasia ........... 329 

Blackfoot Indian (Redskin Type), from a photograph . . . , 338 
Native of Otovalo, Ecuador \ fom w ^ and A ^^ 

Native of Zambisa, Ecuador I , ^ .. ., ., _ .. 

__ . .__ , I photographic album ot Indian 

Dative of Vancouver Island V rf ^^^ ^ ^^ 

Nat,ve of Saqms,!,, Ecuador ^ _ 338,339,3^ 

Paez Indian, of Tacuzo, Colombia/ at) * ooyi ^ 

VT . f r> . . , .,, . ,T, f * -^ T> \) from E. im Thurn's col lec- 
Native of British Guiana True Canb Type) f / , , . , 

,,, . rr-*i /- /* IT \ C ton of photographs in the 
Native of Bntish Guiana (Arawak Type) \ ~ r < f J , 

JL ' J R. Geograph. Soc. 355, 356 

Family Tree of Homo Aaiericanus 361 


Eskimo of Alaska, from A. H. .Keane's Types of Races (see above) . . 363 

Family Tree of Homo Caucasicus 380 

Norwegian (Xanthochroid Type), from A. H. Keane's Types of Races . 381 
A Riff, North Coast Morocco (Berber Type), from a photograph taken in 

Tangier 384 

Berber (West Hamitic Type) "\ , A ^ } ^^^ / 
Somali East Hami.ic Type) \ fro ? A : H " Kea " e S TypK f **"* <"* 
Arab (Semitic Type) J ab Ve) 38 3 , 3 88, 393 

Afghan of the Zerafshan District 1 from the Tashkend Album (see 
Hindu of East Turkistan J above) .... 397, 400 

S^rni Vivekanada (High Caste Hindu Type) i fiom w _ T Stead , s ^ 
M. Khrimian, Catholicos of Armenia (Irano- > // 

Semitic Type) I ^ of R e h gt on S m , 404 

A Tajik of Tashkend (Iranian Type) | from the Tashkend Al- 

A Tajik Woman, E. Turkistan (Iranian Type) J bum (see above) 406, 407 

A Monk of Kikko Monastery, Cyprus (Greek Type), from a photograph 

by Dr F. H. H. Guillemard 408 

A Parsi of Bombay (Iranian Type), from the Congress of Religions (see 

above) 409 

Kabardian of Central Caucasus (Melanochroid Type), from A. H. 

Keane's Types of Races (see above) ....... 416 

Ainu of Urap (Caucasic Type), from R. Hitchcock's The Ainos of 

Japan> Washington, 1892 419 




Definitions Anthropology General and Special Ethnology Ethnography- 
Scope of Ethnology General Nomenclature Definite Terms : Race ; 
Clan; Tribe; Family; Totem; Branch; Stock; TypeIndefinite Terms s 
Division; Section; Group; Horde; Nation; People Example. 

OF the various branches of knowledge, whose subject is man, 
the most comprehensive is ANTHROPOLOGY ! , which 

. Definitions, 

in fact, taken in its broadest sense, embraces all Anthropology 

the others. But as knowledge grows it necessarily Jjjeciai! * nd 
tends to become specialised, and Anthropology, 
the "Science of Man," is now mainly restricted to the study of 
man as a member of the Animal kingdom. It seeks to determine 
the position of the human family in the group of mammals, and 
more particularly to define its relations to the anthropoid apes, 
the nearest genera in the order of primates. Thus special, as 
opposed to general Anthropology, is a science whose object is 
the study of mankind considered as a whole in its separate in- 
dividuality and in its relations to the rest of nature (Paul Brocd). 
But the relations of man to the Anthropoidea are mainly 
physical, and in any case zoological studies take Sco 
little or no account of the mental qualities of special An- 
animals, but only of their bodily structure. Hence thr P olo *3 r - 
it is that special anthropology is concerned above all with the 

1 Gr. &v 0p wires == man; \6yot~ discourse. 


human anatomy, and the anthropologist, as here understood, is 
essentially a comparative anatomist. Again, the Hominida, that 
is, the primary members of the human family, also present 
structural differences, which have to be gauged by comparative 
anatomical studies. Consequently not only man as a whole, but 
also the main divisions of mankind, come to this extent within 
the scope of special anthropology. 

On the other hand these main divisions differ also in their 
Ethnoio mental qualities, and their psychological are at 
least as important as their physical characters. 

Hence special anthropology cannot cover the whole of this field, 
and as it were on the principle of division of labour, hands over 
the detailed study of the Hominidae in all their relations to the 
sister science ETHNOLOGY 1 , which has been aptly defined as that 
branch of general anthropology which deals with the relations of 
the different varieties of mankind to each other (Latham}. Thus 
is clearly seen the essential difference between the two, about 
which confusion still prevails. Anthropology treats its. subject 
l>rimarily from tne P h y sical side ; ethnology treats the same 
subject both from the physical and psychological sides, borrowing 
its anatomical data however from the elder branch. The one is 
more technical and special, the other more all embracing, while 
both must be regarded as mutually complementary. 

Again ethnology differs essentially from ETHNOGRAPiiY 8 with 
which it is also constantly confused, but which 
graphy." * n correct language is rather literature than science. 
It is purely descriptive, dealing with the character- 
istics, usages, social and political condition of peoples irrespec- 
tive of their possible physical relations or affinities 8 . The 

1 Or. 01*05=: race, people. 

2 Gr. ffros and 7pa0i}= description. 

3 Such at least is the general use of these terms amongst English writers, 
and it is desirable that the distinction be maintained, both for the sake of 
clearness, and to avoid the practice of French writers, who almost habitually 
confound ethnology and the synonymous ethnogeny with ethnography, and are 
thus obliged, when precision is essential, to speak of "ethnographic descriptive." 
M. de Rosny, amongst others, gives an unlimited scope to ethnography, 
declaring that it results from "la synthese de toutes les sciences qui ont pour 
but de rechercher la mission de 1'homme et ses destinees"; on which M. J. van 


subjects of ettoo^rajj>Jhy are the^ various groups of peoples 
taken in_dejendentiy one of the other ; the subjects 
f Ethnology are the Jm Ethnology. 

as 2 J5L%5JL cpirckted ^embers of one or m ore 
primordial families. Hence ethnology, like anthropology, neoes- 
Sarily proceeds by the comparative method, co-ordinating its facts 
with a view to determining such general questions as the antiquity 
of man ; monogenism or polygenism ; the geographical centre or 
centres of evolution and dispersion; the number and essential 
characteristics of the fundamental human types ; the absolute ad 
relative value of racial criteria : miscegenation ; the origin and 
evolution of articulate speech and its value as a test of race ; the 
influence of the environment on the evolution of human varieties, 
on their pursuits, temperament, religious views, grades of culture; 
the evolution of the family, clan, tribe and nation. 

In thus defining the scope of ethnology, terms have been 
used which themselves need definition, and all the 
more that the meaning of some, such, for instance, nomenclature.* 
as race, dan, tribe, still gives rise to constant, often 
to angry, discussion, amongst writers on ethnological subjects. It 
is no exaggeration to say that many stout volumes might have 
been spared, had a common understanding prevailed regarding 
the strict sense of the current terminology when the foundations 
of the science were being laid some few decades ago. But in 
speculative branches of research first principles cannot be estab- 
lished by deductive process a priori \ they are rather the natural 
outcome of the inductive method based on cumulative evidence 
a posteriori. 

Ethnological studies have now reached that stage at which it 
seems possible, and therefore desirable, to deter- 
mine the exact meaning of the general terms in terms in corn- 
common use. Such terms as genus, species, variety mon uie ' 
need not here be discussed. They belong to all branches of 
biology, and their meaning is clearly defined in a way that gives 
rise to no misunderstandings. For the ethnologist there is merely 

den Gheyn aptly observes that here "1'ethnographe ne se distingue pas essentielle- 
ment de 1'anthropologiste, de l'arche~ologue, du linguiste, du psychologue" 
(Revue des questions scicnlifiquts, October 1885). 



tjie question whether the Hominidae constitute so many species of 
one genus, or only so many varieties of one species, as will be 
discussed in Chapter VII. 

QLjiiJiiBE^ two Distinct catejjpries, 

one implying affinity or blood relationship of some sort, or at least 
such close resemblance as points at genetic descent from commorf 
ancestry, the other involving no such assumption, vague and inde- 
finite, but therefore in certain cases all the more convenient, and 
indeed indispensable wherever no theories of kinship are involved. 
E*ch has thus its proper place, and it should be specially noticed 
that although the terms of the definite class are 

Definite and . -,.,. , 

indefinite mostly convertible with those of the indefinite, the 

lerms. latter are not to the same extent interchangeable 

with the former, as will presently appear : 


(involving or suggesting the idea 
of kinship). 







Stock ; Stem. 



(indifferent to the idea of kinship). 









After assigning their proper limits to the various branches of 
general Anthropology, Broca sums up with the remark that 
" ^jnographx,studies f. Jgeogjes, eAnolggy_jraces. w Here a sharp 
contrast is drawn between the definite term race 
and the indefinite people, a contrast entirely in 
accordance with the nature of the two subjects. It is obvious 
from the foregoing remarks that ethnography can have nothing to 
do with race as such, for this term, taken in its strictest sense, 
involves common descent from an^ cmginal jstocl^ 
essentially a question of hlpocj- It answers to the breed and strain 
of cattle-farmers and bird-fanciers, and is therefore applicable 


only to groups of individuals sprung, or assumed to be sprung, 
from one and the same original family. 

But mankind has been so long on the earth, and has been 
subject to such endless migrations, displacements and inter* 

minglings of all sorts, that in > the ; ppiniLon_ of mar0 M Spujid 

ej[hnoj^ij^s u few if any pure races now : suryiye. Hence the word 
comes to be used somewhat hypothetically. Certain abstract 
ethnical types are assumed or inferred from a general survey 
of the Hominidae, and the various human groups are classed 
together or discriminated according as they approach or diverge 
from these abstract types. Hence at present race has rathejL.a 
relative than an absolute value, and Topinard regards the word 
as^no more than ** permissive " in _ethnology. He looks upon it 
as synonymous with the natural divisions of the human family, 
however remote the period at which such divisions were consti- 

For Prichard race is a collection of individuals presenting 
more or less common features transmissible J^y RacC| 
succession, in fact, what would now be called "per- specie* and 
manent varieties/' the origin of the characteristics 
themselves being an unsettled question. Pouchet also regards 
race as practically synonymous with species. Hence the word 
will have a different meaning according to the different views 
entertained on the question of the unity or plurality of mankind. 
P^orjtose^who^hold that all the Horn midge form but one spjjcjjgs, 
there can be but one /undameiit^ race, and the current groupings 
are strictly speaking unscientific, however convenient and even 
necessary for the detailed study of the human family. It may be 
Concluded. with Darwin that, at the jnitial^ stage of their 
ragej^ha^ ys^ti^^ojfji^gi 
which tend themselves to become species. Hence i on the as^ 

ysEJ^iJbi^Kli?? ?jJii..^ 

stage, the expressions human varieties and human races are 
practically synonymous, and will be so taken in this worlj. 

Under race come the tribe and the dan* which terms also 
involve kinship even in a narrower sense, being 
properly subdivisions of the race or family groups 
.connected by the ties of blood and recognising a common social 


organisation whether under hereditary or elected chiefs or elders. 
This organisation, which has been diligently studied by Morgan 
and others in recent years, throws much light on the origin of 
human societies; in the hands of these writers it has acquired such 
expansion, that it can no longer be adequately treated within the 
limits of ethnology proper, and it now forms the basis of Sociology 
which has been raised by Mr Herbert Spencer to the rank of a 
separate science. 

Here we are concerned only with the difference between the 

The Clan C ^ an P r P er an( * ^ e tr ^ e ' ^ n tne c l& n System de- 
system of scent was probably at first reckoned only through 
kms p. ^ ema | e ij ne j consequently uterine ties alone 

constituted kinship, the father not being regarded as related even 
to his own children, and not considered as a member of the family, 
as still amongst the Chi (Tshi) people of the Gold Coast and 
elsewhere. In this system all the children bear the clan-name 
transmitted through the mother, and the clan-name thus becomes 
the test of blood-relationship. But the moment descent is 
recognised through the maje line also, as amongst the Yorubas 
of the Slave Coast, the clan system breaks down, and the clan 
merges in the tribe. This point, hitherto one of the puzzles of 
ethnology, has been cleared up by the late Col A. B. Ellis, who 
remarks that " since two persons of the same clan-name may, 
under the clan-system, never marry, it follows that husband and 
wife must be of different clans. Let us say that one is a Dog 
and the other a Leopard. The clan-name is extended to all who 
are of the same blood \ therefore, directly the blood-relationship 
between father and child comes to be acknowledged, the children 
of such a pair as we have supposed, instead of being, as hereto- 
fore, simply Leopards, would be Dog-Leopards, and would 
belong to two clans. They in their turn might marry with 
persons similarly belonging to two clans, say Cat-Snakes, and the 
offsprings of these unions would belong to four clans. The clan- 
system thus becomes altogether unworkable, because, as the 
number of clans is limited and cannot be added to, if the clan- 
name still remained the test of blood-relationship and a bar to 
marriage, the result in a few generations would be that no 
marriages would be possible. Consequently the clan-name ceases 


to be the test of consanguinity, kinship is traced in some other 
way, and the clan-system disappears 1 ." 

It is thus seen that the tribe is not merely a group of clans, 
but that its constitution becomes profoundly modi- 

Th* TriKe 

fied by the gradual substitution of patriarchal for 
matriarchal rights. During this process the exogamous* unions, 
necessary in the clan system to avoid the fatal 
results of too close in-breeding, are continued % l % om 
through force of prescribed usage, the consequence 
being a general weakening of the ties of blood, on which the cfttn 
was exclusively based. The infusion of foreign elements is later 
increased by inter-tribal wars, abduction and the capture of 
women and children. Hence, although the idea of consanguinity 
persists, the tribe, as it expands, depends more and more on 
common social and political institutions, and less on actual 
kinship. Doubtless the foreign elements, entering slowly, are in 
great measure slowly absorbed, so that the physical characters of 
the group are long maintained almost intact ; but the time comes 
when there is no longer any " necessary correlation between the 
social unit which we call a tribe and the physical unit which con- 
stitutes the characteristics of the individuals of a certain region V 
It is, however, to be noticed that during the early period of 
human society the interminglings were necessarily between closely 
allied communities, such as the Italic Latins and Sabines, so that 
the racial integrity would be little affected by such incidents as 
the "rape of the Sabines"; hence the tribe amongst peoples at a 
low grade of culture is still commonly taken as a consanguineous 
group in ethnological writings. Beyond the exact sciences most 
things are relative, and we live in a world of compromise*. 

1 The Yoruba-speaking Peoples of the Slave Coast, 1894, p. 175. 

* Gr. w= outside; y&fju>*= marriage. The convenient term exogamy % first 
proposed by MLennan, implies the custom of seeking a wife outside the tribe, 
and is thus opposed to endogamy (Gr. ft>$ov = within), marriage within the tribe, 
assumed to be a later development. 

8 Dr Franz Boas, Anthropology of the North American Indians, reprinted 
from the Memoirs of the International Congress of Anthropology at Chicago 

(i893), P- 38* 

4 Tribe (Lat. tribw) has been referred to an Aryan word trapti, which 


At the base of the tribe is the family, which is the irreducible 
unit of the clan, but which in Anthropology is also 
ciasTs'terrif taken in a wide sense, though always so as to 
imply consanguinity. No difficulty is presented by 
this larger use of the word, which is applicable to any great 
division comprising a number of more or less closely allied sub- 
groups. Thus all mankind may be regarded as a family forming 
one of the five sub-groups of the anthropoidea. Similarly the 
primary, and even lesser divisions, of the hominidae may be 
spoken of as so many families in reference to the whole group, 
and so on. 

But in its narrower sense no word has given rise to more 
angry discussion than the family, taken as the 
starting point of all human society. It involves 
such questions as original promiscuity, various 
kinds of polyandry, and polygamy as antecedent to monogamy. 
Here it will suffice to state that the assumption of primitive 
promiscuity advocated by so many recent ethno- 
logists is neither necess \ry nor even probable. 
The views of M c Lennan and Morgan, which are supposed to 

Haie's and ^old t ^ ie ^ e ^ at P resent > ^ iave ^> een sharply criticized 
wester- by Horatio Hale 1 , and by Edward Westermarck of 

marc 1 ' Helsingfors (Finland) in his able treatise on The 
Origin of Human Marriage (1890), where the conclusion is 
arrived at that " in all probability there was no stage of human 
development when marriage did not exist, and the father always 
was, as a rule, the protector of his family. Human marriage 

survives in the Gothic thaurp, whence thorp, Ger. dorf. If the equation be 
correct, this word meant originally nothing more than a village group or com- 
munity, whereas the clan was always associated with the idea of kinship ; hence 
Ir. or Gael. clann= offspring, descendants; and kin^ kind are the Sanskrit 
janana, Gr. 7^05, Lat. gens, Old German Chunni, The word mankind itself 
is the Anglo-Saxon mancynn^ implying the ultimate kinship of all the human 

1 In Language as a Test of Mental Capacity -, from the Transactions of the 
Royal Society of Canada (1891), where this ethnologist rejects the cattle- 
herding theory, holding with Darwin that, from the first, man was a pairing 


seems to be an inheritance from some ape-like progenitor * 
(p. 64) l . 

In confirmation of this statement it may be pointed out that 
most if not all of the Simiidse, man's nearest akin, 
live either in family groups or in small parties a^o^sfthc 5 
of several families, and construct arboreal shelters Anthropoid 
where the female and young pass the night. It is PCS ' 
noteworthy that the male gorilla is said to sleep at the foot of the 
tree, while the chimpanzee occupies a forked branch below the 
family resting-place, thus illustrating various stages in the evolution 
of the family life. Some of the New Guinea and Sudanese 
aborigines also build arboreal habitations, in which all the 
members of the family reside, or take refuge from more powerful 
hostile neighbours 9 . The social unit is thus reached by the 
natural process from below, and not with Prof. T. H. Green by 
implication from above. "If asked by what warrant we carry 
back the institution of the family into the life of the most primi- 
tive men, we answer that we carry it back no farther than the 
interest in permanent good. From beings incapable of such an 
interest, even though connected by acts of generation [genetic 
ascent?] with ourselves, we cannot in any intelligible sense have 
been developed 8 ." Those who have studied these questions in 
situ never reason in this way. They know that " primitive men " 
have no thought for "permanent good," though fully aware of the 
present advantages derived from association. It is well under- 
stood even by the Fuegians, who form family groups, but have 
not yet reached the clan state, as shown by the absence of totems, 
the children being named neither from the father's nor from the 
mother's side, but only from the place of birth. Thus all will 
have the same name if born in the same place, and all will have 
different names if born in different places 4 . Here there is no 

1 This work, which is written in sterling English, is of a fundamental 
character, and Reserves to be better known than it appears to be in the English- 
speaking world. But the subject is so vast, that it may almost be said already 
to form a separate branch of the anthropological sciences intermediate between 
ethnology proper and sociology. 

2 Nachtigal, Sdhara und Sudan, II. p* 628. 
s Prolegomena to Ethics, 231. 

4 " I figli non portano il nome dei genitori ma prendono i norni delle localita 


clan, tribe, or government of any kind, but the family exists 

Intimately connected with the primary social division is the 
vexed question of totemism 1 > which Lubbock and 

Totemism. * . 

Spencer trace back to the general practice 9f 
naming persons after plants and especially animals Deer, Bear, 
Turtle, &c. these animals thus in certain cases becoming here- 
ditary family and clan names. But Prof. E. B.Tylor, perhaps the 
first authority on all questions of primitive culture, warns us that 
*%hile granting such a theory affords a rational interpretation of 
_ . . f the obscure facts of totemism, we must treat it as a 

Origin of 

the Totemic theory not vouched for by sufficient evidence, and 
sys em. w ithin our knowledge liable to mislead if pushed to 

extremes" (Prim. Culture, n. p. 215). It is nevertheless now 
commonly assumed with M c Lennan (Fortnightly Rev. 1869 70) 
that all or nearly all peoples have passed through this totem- 
stage of human society. In its present aspect the totemic system 
is thus set forth by Col. Garrick Mallery : " An animal or a plant, 
or sometimes a heavenly body, was mythologically at first, and at 
last sociologically, connected with all persons of a certain stock, 
who believe or once believed, that it was their tutelar god, as 
they bear its name Each clan or gens took as a badge or 
objective totem the representation of the tutelar daimon from 
which it was named. As most Indian tribes were zootheistic, the 
object of their devotion was generally an animal, e.g, an eagle, a 
panther, a buffalo... a snake or a fish, but sometimes was one of 
the winds, a celestial body, or other impressive object or pheno- 

dove nascono...Quindi dieci figli, che nascono in dieci luoghi different!, hanno 
dieci nomi diversi" (Dr Domenico Lovisato, Appunti etnografici...sulla Terra 
del Fuoco, Turin, 1884, p. 34). It is noteworthy that in this lowest known 
form of the family group, it is not the mother but the father that rules, showing 
that matriarchy need not necessarily have preceded patriarchy, as is too readily 
assumed from the study of more advanced social systems wrongly called 

1 From the Algonquian word totem , the proper form of which appears to be 
(item* the distinctive badge carefully guarded by each member of the clan 
(Cuoq, tudc$ philologlques sur qttelques langues sauvages de PAmtrigue, quoted 
by Reclus, XV. p. 480, French ed.). 


menon 1 ." This vie.v may perhaps be accepted, on the con- 
dition of reversing the process of evolution. We are too apt to 
read into the primitive mind our own elevated thoughts on the 
relations between the natural and supernatural orders; else it 
would be seen that the "sociological" must have preceded the 
" mythological" stage. Hence the belief that the totem was "a 
god" or "tutelar daimon," must be regarded, so to say, as an 
afterthought, evolved when the savage mind had become capable 
of such a lofty conception. Palaeolithic man for instance, was 
certainly not a "zootheist," or a "theist" of any kind; yet tlfe 
Dordogne "artists" or the Derbyshire cave-dwellers may well 
have had their family names derived from the surrounding fauna 
and other sources, for, as shown, they were ab initio constituted 
in family groups. The mistake here made is somewhat analogous 
to that of the missionary who deified the augad (totem) of the 
Mabuiag Islanders (Torres Strait), translating the Son of God 
(Mark i. i) by the words Augadau kazi, literally "the Totem's 
Son." The expression remains unintelligible to these Papuans, 
whose totems are still merely family names (crocodile, snake, 
shark, &c.), and have not even reached the rank of demons, 
much less that of the Supreme Being. But the generally received 
theory adapts itself fairly well to the present relatively advanced 
stage of those primitive cultures which have here and there some- 
what dimly grasped the idea of anthropomorphic genii powerful 
for good or evil, mostly evil. 

The terms branch, stock or stem, being borrowed from the 
" family tree," always imply close kinship, and Branch 
their use should give rise to no ambiguity. But stock, 
words of such precision have necessarily less cur- em< 

rency in anthropological than in linguistic studies. This is well 
seen in Powell's classification of the North American Indians 
(Washington, 1891), where the expression "stock language" is 
always intelligible, whereas " stock race " has to be used with 
great reserve. The philologist has no doubt at all about the 
radical difference between, for instance, the Iroquian and the 

1 Picture-writing of the American Indians, in Tenth Annual Report of the 
Btire&u of Ethnology > Washington 1893, p. 388. 


Algonquian stock languages, while the anthropologist scarcely 
admits two distinct. Iroquian and Algonquian physical types. 
Much of the confusion pervading most ethnological treatises 
arises from inattention to this fundamental contrast between 
ethnical and linguistic relations, as will presently be seen. 

Type stands apart from all other general terms in ethnological 

nomenclature. It is not a race, a tribe or a family, 
abstraction. or an y concrete division whatsoever ; but is rather 

in the nature of an abstraction, a model or pattern to 
Which all possible divisions are referable, Originally meaning a 
mould or matrix, or rather a casting from a mould 1 , it is taken as 
a summary of all the characters assumed to be proper to a given 
class or group. Thus type becomes the standard by which we 
measure the relative position of individuals in a group. But in 
practice no individual exists, or ever did exist, who is entirely 
conformable to any given standard. Hence type necessarily 
resolves itself into a question of averages ; individuals possessing 
most of the characters peculiar to a group are said to be typical 
members of that group, and even this only in a relative sense. 
They approximate nearer than other members to the ideal, but 
none absolutely reach it. There is, for instance, no perfect 
embodiment of the Caucasic or of the Mongolic type, and it has 
been well remarked that "a large proportion of mankind is made 
up, not of extreme or typical, but of more or less generalised or 
intermediate forms 2 ." 

Exaggerated specimens, hypertypes, as they are called, do 
however occur, but only in one or two respects; such are the 
Fijian Kai Colos, who are said to be "hypertypical Melanesians," 
because of the excessive dolichocephaly of their crania (Flower). 
But it would be rash to assert that these aborigines, of whom little 
is otherwise known, are even typical Melanesians in every respect. 
Other forms of the word, such as proto-type, sub-type, etc., explain 
themselves. But it should be noted that in its simplest form 

type is also occasionally used in a concrete sense, as 

when we say that the Ba-twa are a type of the 
. African Negritoes, meaning that they are typical 

1 Gr. Ti$7rrw, to strike, hence a stamp or distinguishing mark. 
* Flower and Lydekker, Mammals living and extinct, p. 744. 


negritoes. This use of the word, however, often gives rise to 
misunderstandings, which, as observed by M. Sanson, might be 
avoided by using the more definite expressions "racial type," 
"specific type," and so on 1 . 

Most of the Indefinite terms also explain themselves, and with 
ordinary care can scarcely lead to any misunder- 
standings. The chief point to observe is that the ni h t c rms Cfi ~ 
definites and indefinites are not necessarily inter- 
changeable, because the latter do not connote or involve the 
attributes of the former. Thus a branch is always a division^ 
section or group ; but divisions, sections or groups need not be 
branches. J*he Melanesians are a branch, section or division of 
the Negro stock ; they are also a section or division, but not a 
branch, of the Oceanic peoples, who do not form a family group. 
The Kipchaks may be called a tribe or a horde indifferently of the 
Usbeg branch of the Turki race ; but they are a horde only 
of the Mongolo-Turki nomads, their affinities being with the 
Turki, not with the Mongol division. 

Horde from yurt> urdu, a tent, then a group of tents, camp, host, 
army, differs from tribe, in that it implies no kinship, 
but only a group of nomads brought together for 
predatory or other purposes. Many of the " Tatar hordes " were 
not Tatar (Mongols) at all, but of Turki stock. The Sudanese 
Sofas, are not a " tribe," as they have been described, but a horde, 
a band of riff-raff from all the surrounding tribes (Mandingans, 
Fulahs, Bambaras and others), brought together by Samory to war 
against the French in the Upper Niger basin, and to raid the land 
for slaves and plunder (1893-95). This is the obvious historic and 
general ethnological use of the term horde, which has to be made 
the starting point of human society only by those writers who with 
M c Leftnan evolve order out of "promiscuity." "If we may properly 
dismiss the term family as a scientific appellation for the earliest 
group -of human beings, and if we may consistently call it the 
horde, borrowing the term from M c Lennan, we shall, at least, 

1 Bui. de la Sec. d*Anthrop. June Oct. 1889, p. 400. AH will agree with 
this naturalist that **il y aurait a vantage a ne pas se servir, dans le langage 
anthropologique, d'expressions vagues qui ne servent qu'a obscurcir les ide'es " 

ttb,). . . . . 


be clearing the way to prevent a misconception from a confusion 
in terminology 1 . 1 ' Here horde is taken as practically synonymous 
with her^ and " confusion " would be better avoided by frankly 
making the substitution. Then horde will recover its historic 
significance, and M c Lennan's "Theory of the Primitive Human 
Herd** will at least have the merit of putting the point at issue 
in intelligible language. This "cattle- herd" theory is a pure 
assumption, an unscientific deduction, based on blank ignorance 
of facts that can never be known until the social life of eolithic 
Eian is recovered. The nearest analogous case is the social life of 
the higher apes, who, as seen, do not herd together, but live in 
family groups. But the less is known about these relations, the 
thicker the tomes devoted to their exposition. 

Through the horde the tribe expands into the nation wt\& people, 
where the idea of race or kinship is destroyed by 

Nation. . , rTVI r . , 

universal mixture. The nation comprises all the 
inhabitants of a given region subject long enough to one political 
system to have acquired a certain outward uniformity, a common 
standard of social usages, interests, aspirations, generally also 
language, literature and religion. But although not involving 
common origin, it tends towards ethnical uniformity or unity, by 
the gradual fusion of diverse elements in a uniform type. Some 
nations, such as the Swedes, have in great measure acquired such 
uniformity, and with them race and nation become practically 

convertible terms. People is a still more elastic 

People. . . , . ... 

expression, and may be taken to comprise in the 
singular all the uncombined sections of the nation, in the plural 
an aggregate of nations remotely connected by vague traditions, 
allied languages and especially a common social culture. Thus we 
speak of the " Hungarian people/' an expression which includes 
the Ugrian Magyars, the German Transylvanians, several Slav 
groups, the Rumanians of Latin speech, and others scarcely yet 
merged in a common nationality although living under a common 
political administration. So also the wider expression " European 
peoples" embraces many nations mostly of Aryan speech and 

1 G. L. Com me, Jour. Anthrop. 2wt* 1888, p. 119. 


The foregoing remarks may be illustrated by the following 
example, in which the definite and indefinite terms B Xftmp i e H. 
are in italics : Dr Oronhyatekha, who visited England iwtrting the 

use of the defi- 

in 1894, is a typical member of the Turtle dan of nites and in- 
the Mohawk tribe, who are a branch of the Iroquois deflnlte * 
nation. The Iroquois themselves, who speak a stock language, 
form an important section of the American variety of the Homi- 
nidse, that is, of the human family in the sub-order Anthropoidea of 
the order Primates, 



Man's Place in the Animal Kingdom The Primates Old Divisions: Quad- 
rumana and Bimana New Divisions: Lemuroidea and Anthropoidea 
The five families of the Anthropoidea Their range in time and space 
Diagram of the Anthropoid families Relations of the family Hominidae to 
the family Simiidse Comparative Table of the Simiidae and Homim'dse : 
Gibbon; Orang; Gorilla; Chimpanzee; Dryopithecus ; Hominidae 
Points of resemblance to and difference from the Simiidae Origin of 
Man by Creation or Evolution Creation Theory inadequate Evolution 
Theory adequate Natural and Supernatural views reconciled Difficulties 
of the progressive evolutionist theory Views of de Quatrefages, de 
Mortillet and Sergi The Castenedolo Man Sergi's Tertiary Hominidse 
Quaternary Man Cannstadt Man rejected Neanderthal affirmed The 
Quaternary Hominidae Kollmann's Dauertypus Persistence of primitive 
types Views of French, English and American Anthropologists Diffi- 
culties of the Dauertypus theory Analogy of the Equidse Their evolution 
Sergi's Tertiary Hominidae rejected Persistence of, and Reversals to, 
primitive types reconciled with evolutionary teachings Comparative 
Diagrams of Pleistocene Hominidse and Equidae Broad stages of physical 
evolution from a postulated Anthropoid Miocene precursor. 

AT the end of the foregoing Chapter man's position in the 
Man's lace aiuma l kingdom was determined from the purely 
in the animal zoological standpoint. That he is an animal, and 
mg om. ag guc k mu st b e related to other animals, is no 

discovery of modern science. Even the schoolmen defined him 
as animal rationale, a definition which the ethnologist may accept 
without hesitation as at least partly true. What modern science 
has done is to give precision and completeness to this definition, 
by fixing the place of man as an animal in the class of mammals, 
and by separating him, mainly in virtue of his exclusive posses- 
sion of articulate speech, from other animals to whom the reason- 
ing faculty can scarcely be denied. Man will accordingly here be 


considered as a rational animal possessing the faculty of articulate 

Zoologists distinguish in the Class Mammals the large and 
widespread group of Apes and Half- Apes (Lemurs), 
which in all modern systems constitute the inde- 
pe'ndent order Primates, so named by Linne' because viewed as 
a whole they are the chief or most highly specialized o .. .. . 
members of the class. Cuvier made two divisions sions. 
of the Primates, the Quadrumana or "four-handed," Quadrumana 
comprising all the Apes and Half- Apes, whose four and Btmana> * 
extremities are more or less prehensile, and the Bimana or 
"two-handed," comprising the single genus Homo, that is, man 
alone, whose two anterior extremities are prehensile (true hands), 
while the two hinder are true feet But these terms, after playing 
a large part in anthropological writings, have fallen into disuse, 
being not only unsatisfactory but even misleading, " Anatomic- 
ally the foot of apes agrees far more with the foot of man than 
with his hand, and similarly the ape's hand resembles man's hand 
and differs from his feet. Even estimated physiologically, or 
according to use, the hand throughout the whole order [Primates] 
remains the prehensile organ par excellence, while the predominant 
function of the foot, however prehensile it be, is constantly loco- 
motive. Therefore the term Quadrumana is apt to be misleading, 
since anatomically both apes and man have two hands and a pair 
of feet 1 ." 

Excluding the Bats, which had been classed by Linne with 
the Primates, recent systematists split the Order into 
the two suborders Lemuroidea and Anthropoidea, 
and subdivide the Anthropoidea 3 . that is, the "man- orders 

,., r ... _ - ... / i i , TT Lrcmuroidca 

like forms into five families, of which the Homi- andAnthro- 
nidce constitute the fifth. Of the other families two, P ldca * 

1 St George Mivart, Man and Apes, p, 88. 

8 Gr. "Av6pwiros and etSoy = form, shape ; hence in the shape, form or like- 
ness of man. The term efSos as a suffix has a wide application in ethnology, in 
which it serves to distinguish between groups conformable to and more or less 
divergent from a given type. Thus : the Ashanti are true Negroes, whereas the 
Zulu-Kafirs are rather Negroid. So Mongoloid, Caucasoid, Australoid and 
so on. 

K. 2 


The five Hapalida (Marmosets) and Cebidce (American 

families of the , \ i i A j A XT 

Anthropoidea. monkeys) are exclusively confined to the New 
Family World, while the two others, Cercopithetida (baboons, 

(hMiest a ) macaques and other apes generally long-tailed), and 
their ran e SimiidcR (Gibbon, Orang-utan, Anthropopithecus or 
in time and Chimpanzee, and Gorilla, all tailless) are confined 
space. to t ^ e tr0 pi ca i an( .[ temperate' regions of the Old 

World. The range of the Simiidse is still further restricted at 
present to the tropics, Gibbon and Orang to Indo -China and 
Malaysia, Chimpanzee and Gorilla to equatorial Africa. But in 
former times they extended far beyond these limits, the fossil 
remains of a true Chimpanzee Pal&opithecus or Anthropopithecus 
sivalensis) having been found in the pliocene of the Panjab, an 
extinct Gibbon (Pliopithecus) in the Middle Miocene of France, 
and in the same geological formation a somewhat generalized 
Simian (Dryopithecus) approximating nearest to the Chimpanzee, 
but also showing affinities to the Gorilla, and even to the Cerco- 
pithecidse. But it should be remembered that in the Miocene, 
and generally before the first ice-age of which there is any 
evidence, the tropical zone extended far beyond its present 
limits, and certainly comprised all the regions where these 
fossil apes have been found (see p. 22, 24). Hence it is im- 
possible to infer from their remains that the Simiidae at any 
time lived in colder climates than that of their present equatorial 

It is important to note that man has points of contact with all 
these genera of the Simiidae, as well as with the other families of 
the suborder 1 , and further that "the differences between man and 
the Anthropoid apes are really not so marked as those which 
Diagram of separate the latter from the American monkeys 2 ." 
P os to n m the suborder is placed in a clear 

Anthropoid light in the accompanying diagram of the five 
primates. families, where the distance between the Hominidae 

1 So marked are the links with the half-apes that some zoologists trace 
man directly from the suborder Lemurs without passing through the line of 
anthropoids at all. This view, however, seems somewhat paradoxical, and 
.does not appear to have been taken seriously by specialists. 

8 Flower and Lydekker, op. (it. p. 740. 



and the Old World anthropoids is less than that between the 
latter and the two New World groups : 







It is popularly supposed that, according to evolutib&ary 
teachings, man is "descended" from the Gorilla, or Relations of 
the Chimpanzee, or from some other member of the man to the 
Simiidae, his nearest congeners. But no sane imu *' 
evolutionist holds such a doctrine, and from this diagram it is 
seen that the ascent of man is taken to be in an independent 
line from some long extinct generalised form, from which the 
other branches also spring in independent lines. All have some 
features in common, while each presents some special characters, 
the source of which is to be sought partly perhaps in a common 
precursor, but mainly in their independent development along 
divergent lines of growth. At the same time the points of 
resemblance between the Hominidae and the Simiidae are far 
more numerous than between the Hominidae and any other 
group, from which it may be inferred that the divergence of 
the higher groups really took place in the sequence indicated 

2 2 


in the diagram. Hence for the ethnologist the study of man 
Comparative fro m the physical side is naturally confined to his 
five^eneraof position in respect of the higher apes. These 
theSmuidee alone are consequently included in the subjoined 
family* * Table, in which those points only are indicated 

Hominidae. t j iat nave a more or less direct bearing on the 
relations of the Hominidae to the Simiidae 1 . 

GIBBON (Hylobates)\ Range, Indo-China and Malay Archi- 
pelago; Species, Siamang (Hylobates syndactylus\ of Sumatra, 
fergest (3 feet high), and specially remarkable for its well-de- 
veloped chin, in this and its wide chest approximating nearer to 
man than any other ape; Hoolock (H. hoolock) of South Assam and 
Upper Irawady Basin, with distinctly aquiline nose ; White-handed 
Gibbon (If. lar) of Tenasserim and Malaysia; Dun-coloured 
Gibbon (H. entelloides) of Malaysia; Tufted 
Gibbon (ff. pileatus) of Siam and Camboja; the 
extinct Pliopithecus of the Middle Miocene of France. The 
Gibbons are the only apes that habitually walk erect, with a 
quick waddling gait, resting on the hind feet alone with sole 
planted flat on the ground, great toe wide apart, and the dis- 
proportionately long arms held upwards, sometimes horizontally. 
Voice much like man's at a distance a peculiar wailing note and 
a double call (koo-lock). Dorsal vertebrae 12 to 14; ribs 7 to 8 
on each side with angles more marked than in any of the other 
genera except man; slender figure. 

ORANG-UTAN (Simid) : Only two known species of which S. 
satyrus is confined to Borneo and Sumatra; red-haired; male 4ft. 
4 in. high, bulky body, extremely short legs, and long arms reach- 
ing when erect down to the ankles, walks slowly and deliberately, 
resting on the knuckles of the hands and the outer sides of the 
feet, the soles being turned mainly inwards; habits arboreal; 
nests of boughs and leaves; food exclusively vegetable. The 
reddish-brown hair covers the whole body, and in adult males 
forms a well-developed beard; thumb and great toe very short; 
dorsal vertebrae 12 as in man : 12 pair of ribs; brain 


in its convolutions more human -like than any 
1 Data mainly from Flower and Lydekker, op, cit. 






other ape's; skull highly vaulted and brachycephalic; superciliary 
ridges moderately developed and not prominent as in Gorilla; 
canine teeth of male very large; lips thicker than in any other ape. 
Remains of a species of Orang (a broken canine) have been 
found in the Lower Pliocene of the Siwalik Hills, showing former 
range northwards to the Himalayas. 

GORILLA (Gorilla) only one known species (G. savagei) con- 
fined to West Equatorial Africa, between the Cameroons and Lower 
Congo river ; massive body and limbs thickly covered with blackish 
hair; larger in bulk than man, but owing to the short legs never over 
5ft. 6in. high ; in upright position arms reach to middle of lower 

leg; very short thumb, great toe relatively longer; 

digits of hands and feet partly united by integu- 
ment; skull dolichocephalic, with enormous superciliary arches, 
giving the gorilla its peculiarly ferocious aspect ; canines of male 
very large and inclined outwards in both jaws ; ear small but well 
developed, with a rudiment of the human lobe; orbits also more 
human in shape than in other apes; dorsal vertebrae 13, as in 
chimpanzee ; wrist lacks the os centrale, as in chimpanzee and 
man ; brain convoluted like orang's, but in some other respects 
comes nearer to the human type, from which it differs only in its 
inferior size and weight and in the more symmetrically convoluted 
cerebrum, which is less complicated with secondary and tertiary 
convolutions. The gorilla walks with backs of closed hands and 
fiat soles planted on the ground; voice a deep guttural sound, 
varying from a grunt to a roar ; lives in family groups, female and 
young passing the night in the trees, at foot of which the male is 
said to sleep ; he is much larger than his mate, and tends to turn 
greyish in old age. 

CHIMPANZEE (Anthropopithecus) : Two known species (A. tro- 
glodytes and A. calvus), ranging over inter-tropical Africa, besides 
the extinct A. sivalcnsis of the Pliocene of Northern India ; both 
sexes much alike, the chief difference being the larger canines of 
the male : extreme height 5 feet ; colour blacker than in gorilla 

and ears relatively larger, but arms shorter, reaching 

Chimpanzee. 1/^1- 

only a little below the knee ; the Chimpanzee 
comes nearest to man in this as in some other respects, such as 
the better developed thumb and great toe, presence of whiskers, 





eye-brows and lashes, more rounded and far less rugged coronal 
region of the skull, relatively much smaller canines, and in the 
dentition generally ; is also more gentle, intelligent and social, 
living not only in family groups, but even in parties of several 
families; builds arboreal shelters for the female and young, the 
male sleeping lower down, according to some observers. 

DRYOPITHECUS, an extinct anthropoid from the Middle 
Miocene of France, same size as and apparently nearest allied to, 
Chimpanzee ; figures largely in ethnological writings, being taken 
Vy certain theorists as the precursor of man in Europe. But Dryo- 
pithecus appears to represent a more generalised member of the 
family, that is, an earlier and simpler type from 
ryopi ecus, w kj cn ^ Q our living genera of Simiidae and the 
family Hominidae may have ascended. Consequently Dryopithecus 
must have been farthest removed from man in time, and must 
have possessed fewer properties of a specially human character. 
Between it and man many more gaps would have to be bridged 
than, for instance, between Chimpanzee and man. But, as 
above shown, man cannot be derived directly from Chimpanzee ; 
how much less from Dryopithecus ! So much may be inferred 
from what little is known of this fossil, which is represented 
only by a single species, whose dentition stands at a stage of 
evolution intermediate between the Cercopithecidae and Gorilla, 
consequently lower than that of all the Simiidae. "A gradual 
transition in the form of the mandible may, indeed, be traced 
from Dryopithecus, through Gorilla, to Anthropopithecus " (Flower 
and Lydekker); and this is all that can be said for the "human 
characters" of Dryopithecus, on which so many fanciful theories 
have been built. How little it is may be seen from the fact that, 
although the mandible (lower jaw) of Anthropopithecus (Chim- 
panzee) approximates in some respects to the human, it does so 
in less degree than that of the gibbon, which on the whole is the 
least human of all the Simiidae. In any case it would have to be 
shown that the later evolution of Dryopithecus proceeded rather 
in the direction of Hominidae than of Simiidae ; there is of course 
no evidence of this, although in Prof. Sollas' diagram (Nature, 
Dec. I9> 1895) a suggested line of human ascent diverges from the 
stem either of Pliopithecus or of Dryopithecus. 


HOMINID^E. (Linne's Genus Homo), forming one species, with 
four primary varieties, all connected by endless inter- 
mediate forms ; spread over most of the habitable ni( J|? e H ^ IBl " 
world from Pleistocene times ; in this work the four 
varieties are named : Homo dEtthiopicus (Negro), most of Africa and 
Australasia ; Homo Mongolicus, most of Asia and Malaysia, parts 
of Europe ] Homo Caucasicus, most of Europe, parts of Asia and 
Africa originally; Homo Americanu$> all the New World. 

Main points of resemblance with the Simiidae in physical 
structure : i. The general anatomical structure, the 
framework of all being cast on the same lines, so w fn "i^ th< 
that all may be conceived as merging or tending to Simiid in 

. , , , , , , structure. 

merge one in the other, although at present the 
primary groups are not united by any intermediate forms. 2. The 
complete disappearance of the tail in all, nothing remaining 
except a few caudal vertebrae, invisible in the living subject ; hence 
the folly of the quest for " tailed men," all reports regarding whom 
(Borneo, Equatorial Africa) may be dismissed as fabulous, If no 
species even of genus Gibbon can show any external trace of this 
appendix, it cannot be expected in any variety, however low, of 
the Hominidae, by far the most specialised group in the suborder. 
3. The two anterior and the two hinder extremities fairly well 
developed throughout as true hands and feet respectively. 4, The 
dentition, which as regards the number and sequence of the teeth 
is the same in all the Old World Anthropoidea as in man. 5. 
Ear, universally well developed, but lobeless in the Apes, Gorilla 
alone showing a rudiment of the human lobule. 6. Brain, which 
in form and general structure is much the same in man and Apes. 
7. Hyoid bone. 8. Liver. 9. Ctecum, all identical. 

Main points of difference from the Simiidae: i. Brain, 
absolutely as well as relatively much smaller in all 

, . , , , ^ Points of 

apes than in man ; highest cranial capacity of difference from 
Orang and Chimpanzee, which in this respect theSinwidae - 
approximate nearest to the human, 26 and 27 \ cubic inches 
respectively; lowest normal in man 55. 2. Brain-case, much 
larger in man relatively to the facial part of the skull. 3. Vertebral 
Column, completely adapted in man alone to the erect position, 
being curved doubly so as to perfectly balance the head. 4. Legs* 


much longer in men than in apes relatively to the arms. 5. Great 
toe, longer in man, and not opposable to the other digits. 
Although this is one of the most marked differences between the 
two families, it is not a strong zoological character, since it 
depends on a slight change in the form of a single bone (the 
entocuneiform) ; but great toe not opposable in the human foetus, 
as has been asserted 1 . 6, Nose, well developed in man, very 
slightly prominent in the Apes, except the Hoolock Gibbon, 
in which it is aquiline. 7. Dentition , forms in man alone 
&n uninterrupted series of horse-shoe shape, without prominent 
canine teeth ; but the canines of the lower and higher races differ, 
the crown of the former being larger relatively to the neck, and 
terminating, like those of the Apes, in a sharp point, usually much 
worn (. Regnault). 8. Chin^ developed in man alone, though 
in one Gibbon (Siamang) the union of the mandibles forms a 
slight projection like the human chin. On the other hand "there 
is no chin in the jaw of the Cannstadt [Spy] race, and the large 
angle approaches without nearly equalling that of the anthropoids" 
(Cope, Genealogy of Man, Amer. Naturalist, April 1893, p. 330). 
9. Hair, covers the whole body in all the Anthropoids, restricted 
mainly to the scalp and parts of the face in man ; the human hair 
is woolly, or quasi-woolly in Homo ^Ethiopicus, never in any of 
the true apes, or even of the half-apes, except the Woolly Lemur 
(Avahis laniger) of Madagascar. Greyness with years is also a 
human feature, although the tendency has been observed in 
Gorilla. Traditional reports regarding "hairy men" have been 
verified only in the case of the Ainu people of North Japan, with 
whom the character is certainly racial. 10. Voice, inarticulate 
in all anthropoids; articulate in all the hominidse, who conse- 
quently are alone endowed with the gift of speech. This, with 
the associated intellectual qualities, separates man altogether 
from his anthropoid congeners, and consequently from all other 

1 "Jusqu'au troisieme tnois de la vie intra-uterine 1'articulation du gros 
orteil du foetus humain est oblique, exactement comme chez les singes; c'est 
seulement au quatrieme mois qu'on la trouve transformed" (Salmon, Races 
Humaines JPrMiistortques, p. 4). Yet the great toe is somewhat opposable 
among the Annamese, who from this circumstance have always been known to 
the Chinese as Giao-Chi or "Crosstoes." 










members of the animal kingdom, with which his connection thus 
appears to be mainly physical. As an animal he is a member of 
the Anthropoidea; as a speaking and reasoning animal he stands 
apart and alone in a separate category, for "the essential attributes 
which distinguish man, and give him a perfectly isolated position 
among living creatures, are not to be found in his bodily structure" 
(Flower and Lydekker, p. 739). 

The general relations of man to the Simiidae, as seen in the 
skeleton, are shown in the accompanying diagram, reduced from 
Juxley's Man's Place in Nature. 

From the data supplied by this Table, one important point 

comes out very clearly that man cannot possibly 

man* cither by ascen( * 1 directly from any of the living anthropoid 

evolution or apes. He has some physical features in common 

creation. . . 

with each, some different from each, some, and 
these not the least important, entirely peculiar to himself. Con- 
sequently, even on the physical side he stands somewhat apart 
from, but yet so near to, all, that his origin can be explained only 
by a process of natural evolution from a common precursor, or 
else by direct creation ; there is absolutely no other alternative. 

If what may be called the deus ex wachina view be 
theory tum accepted, then the gate to further inquiry is closed, 

ethnology becomes ethnography, and we revert to 
the stage at which the question stood in the uncritical times when 
dogmatism usurped the chair of science. Only dogmatism would 
still have to grapple with many new and hopeless difficulties, 
resulting from the cumulation of a multiplicity of fresh facts, which 
point unmistakeably at natural processes of development. Such, 
for instance, are the caudal vertebrae and other rudimentary * or 
atrophied 8 organs which man has in common with the anthro- 

1 Obviously ascend^ not desetnd; so Broca : '* Quant moi, jc trouve plus 
de gloire a monter qu'a descendre, et si j'admettais Intervention des impres- 
sions sentimentales dans les sciences, je dirais que j'aimerais mieux etre un 
singe perfectionne qu'un Adam de'gene're'" (Mtmoires d* Anthropologie^ m. 
p. 146). 

a Owing to their careless use even by scientific biologists, some confusion 
prevails regarding the meaning of these two terms, which are not by any means 
synonymous although constantly taken as such. A rudimentary organ is one 


poids, and which, being useless, are inexplicable on the assumption 
of creation by infinite wisdom, but quite intelligible by the theory 
of evolution. " In order to understand the existence of rudi- 
mentary organs we have only to suppose that a former progenitor 
possessed the parts in question in a perfect state, and that under 
changed habits of life they became greatly reduced 1 /' 

But the supernatural view can in no way get rid of evolution, 
which is indispensable to any theory that attempts 
to account for many patent facts in the natural 
history of the Hominidae. It is not, for instance, 

pretended that all the Hominidae were independently created, but 
one only. Consequently the transition from, say, the Homo 
Caucasicus (if he was the starting-point) to the Homo ^Ethiopicus, 
must have been effected by some natural evolutionary process ; 
from this there is no escape for the creationist. Now the typical 
white man differs enormously from the typical Negrito, so much 
so that they would have to be regarded as separate species but for 
the intermediate forms in actual existence. Here then we have in 
any case a range of evolution scarcely less 5 than that which is 
covered by the transition from Gibbon to Orang or Chimpanzee. 
The difference is obviously one of degree only, and not of kind, so 
far as regards physical structure. 

Creation being thus useless unless supplemented by evolution, 
it may be dropped, or reserved for such points, if Evolution 
any there be, which cannot be explained by the theory ade- 
natural process. This process is adequate for the 
early or initial stages of development ; it may therefore be safely 

so to say, beginning its life history, as for instance, the elementary chin and ear- 
lobe in some anthropoids, which are more fully developed in man. An atrophied 
organ is one, on the contrary, which has run its course, completed its life 
history, leaving only a trace of its former existence, as, for instance, the caudal 
vertebrae, remnants of a tail in the anthropoids and in man. Rudiment is the 
Lat. rudinicntum, front rut/is, rough, unfinished ; atrophy is the Gr. ArpoQla, a 
wasting away through lack of nourishment, from d negative prefix and r^^w, 
to nourish. 

1 Darwin, Descent of Man , p. 15 of 1885 ed, 

2 Huxley implies that it is even more, for he declares that the gulf between 
civilised and savage man is wider than that between tlie savage and the highest 
apes. Man's Place in Nature, p. 78. 


concluded that primitive man, whether it be Haeckel's Homo 

primigenius alalus, or any other postulated form, was not created 

Creation not ^ut evo ^ ve< ^ ^ nee< ^ scarcely be added that 

excluded abso- creation, as possibly standing behind all living 

utey * organisms, is not excluded absolutely, but only 

relatively to the Hominidae, with whom alone ethnology is con- 


This attitude towards creative agency leaves science unshackled, 
and gives a free hand to the biologist within the limits of the 
existing order, without prejudice to dogmatic pre- 
conceptions, without offence to extremists on either 

views recon- s i(j e . it has the further advantage of obviating the 
irreverent and unorthodox 1 introduction of the Ens 
Snpremum, with Cuvier and other supernaturalists, to account for 
every successive change in the animal series, or else of needlessly 
and rashly abolishing the Ens Supremum altogether, with the 
presumptuous modern materialistic school. Natural philosophy is 
not called upon to solve, or indeed to deal at all with, these 
transcendental questions, which may well be left to metaphysicians 
and theologians. It has certainly no right to dogmatise over 
subjects beyond its sphere, about which it can never know any- 
thing, although it may justly claim the right to dispense with the 
aid of the creative force within the limits of legitimate speculation. 
In this middle course would seem to lie the true ultimate " recon- 
ciliation of Science and Religion*." 

t) because such implied bungling and tentative efforts to arrive 
at more perfect types of organic life are derogatory to Infinite Wisdom ; unor- 
thodox t because multi -creation is not warranted by, but opposed to, Scripture, 
which speaks only of three creative acts within the biological horizon two for 
the vegetable and animal kingdoms, and one for man. In the evil days of 
rampant sacerdotalism Cuvier must have shared the stake with Giordano 

2 So Lamarck: "Sans doute, rien n'existe que par la volonte du Sublime 
Auteur de toutes choses. Mais pouvons-nous lui assigner des regies dans 
i'execution de sa volonte* et fixer le mode qu'il a suivi a cet e"gard ? Sa puis- 
sance intinie n'a-t-elle pu creer un ordre de choses qui donnat success! vement 
['existence a tout ce que nous voyons comme a tout ce qui existe et que nous ne 
connaissons pas?... Res pedant done les decrets de cette sagesse infinie, je me 
renferme dans les bornes d'un simple observateur de la nature " (Philosophic 
t % I. ch. in.). 


Man being thus evolved, evolutionists themselves are con- 
fronted with a difficulty which they cannot shrink 
from discussing. It is obvious from his physical 

relations to the Simiidae that the beginning of his 8ive evolution. 
evolution must date from the incalculably remote 
miocene times, when Dryopithecus, and no doubt other fairly 
developed anthropoid forms, had already made their appearance ; 
for, as seen, from none of these can man be derived Where then 
does he part company with all other branches of the suborder, and 
enter on his own proper upward development ? Here the absence 
of intermediate links, owing to the necessarily imperfect state 
of the palaeontological record, seems to bar further inquiry except 
of a purely speculative character. De Quatrefages, 
in fact, and his followers give up the problem, and 
while claiming to be evolutionists to a certain 
extent, they deny that the gulf between animality and humanity, 
as well as between the other animal orders, can be bridged over 
by any process of progressive evolution. The position is exactly 
analogous to that of certain philologists, who admit evolution 
within each morphological order of speech, but not between the 
several orders themselves. 

Others with Mortillet (Prehisforique^ p. 104) claim as the 
precursor of man a tertiary anthropoid, not how- Mortiliet 
ever the Dryopithecus, so that the diverging point 
from the pithecan stem towards quaternary man, admitted by 
all, need not be dated farther back than tertiary times. The 
tertiary precursor himself is assumed to have been near enough to 
a true man to have made implements resembling those found at 
Thenay in France and at Otta in Portugal, and consequently 
superior to those of the recently extinct Tasmanians, who were 
certainly true men. The man of Otta, however, is not yet proven. 
Such was the opinion of Evans, Virchow and de Quatrefages who 
took part in the discussion on the subject, when the Otta flints were 
produced at Lisbon by their finder, Carlos Ribeiro. But on that 
occasion both Virchow and de Quatrefages treated it as a local 
question, and declared their belief in tertiary man on other 
grounds: "Je crois a son existence, mais pour d'autres raisotis" 


Sergi, a leading Italian anthropologist, rejects both of these 
views; that of de Quatrefages as opposed to 
numerous morphological facts in the animal and 
human groups, and as reviving the exploded theory of fixed 
species, and leading to the inconsistency of A. R. Wallace, who, 
although one of the originators of the doctrine of selection, ex- 
cludes man as specially created 1 ; that of Mortillet because no 
postulated Dryopithecus, or any analogous type, can develop 
into the human form. Ail alike are already too highly specialised 
in other lines of development. Sergi holds not only that the 
Thenay objects, but even the fossil remains from 
ne " tne tertiary deposits of Castenedolo in the Brescia 
district, found in situ by Ragazzoni in 1860, and 
allowed by de Quatrefages 8 to belong to the tertiary epoch, 
represent fully developed human beings, so that man would have 
already appeared in that epoch, endowed with all the characters 
by which he is at present distinguished from the lower mammals. 
But Sergi, who may be taken as the most powerful champion 
ser i's ter- ^ " tert i ar y man," goes much further than this, and 
tiary Homi- expresses his conviction not only that man, but the 
01 *' Hominidre themselves, whom he regards not as mere 

varieties but as true species, had already been differentiated as 
genus homo in pliocene times, from which period the different 
species have, like the anthropoids, constantly maintained their 
distinctive characters. Here are, therefore, not merely one but 
several human prototypes, which would have persisted with little 
change (a mere "series of ethnical modifications") from the pro- 
digiously remote early tertiary period down to the present day 8 . 

1 Contributions to the 7'fieory of Natural Selection, 1878. 

* * * A coup stir, si elle [la d^couverte] avait &e faite dans un terrain quater- 
naire, personne n'en aurait contest^ la ralite." Races Humaines* I. p. 100. 
The remains comprised a man, a woman, and two children unearthed under 
conditions afterwards verified by Sergi himself. 

8 ** lo sono convinto che le specie umane si sono formate molto probabil- 
mente nei primi periodi terziari, e sono apparse come genus homo nel plioceno, 
dal qual tempo, come le antropomorfe, ban conservati costantemente i loro 
caratteri. Dopo ci6, una serie di modificazioni, cbe noi chiameremo etniche, 
si sono prodotte per diverse cause &c." (JEvolusione umana, 1888, p. n.) 


Whatever be said of this view, there is no doubt at all not 
only that man, but even several varieties of Homi- 
nidae, had already appeared in quaternary times m j^* tern * ry 
in Europe, Africa, Java, South America and else- 
where. Attempts have been made to discredit some of the 
evidence bearing on this point, and no conclusions Evidence of 
can certainly be drawn from the skull found at cnn*udt 
Cannstadt nearly two hundred years ago, and some- * U rc cc e * 
what hastily taken as representing a palaeolithic " Cannstadt race." 
It is even doubtful whether this skull, now preserved in Stuttgart, 
is the one actually found, not in a quaternary bed as was said, 
but associated with some potsherds in the talus or rainwash at 
the foot of the cliff, on which is a modern cemetery. It may be 
quite recent and probably pathological, or else a Evidence of 
reversal to a palaeolithic dolichocephalic type with Neanderthal 

. , , ^ , . - , . , . skull affirmed. 

simian characters, the existence of which is now 
established by overwhelming evidence. This type upon the 
osteological peculiarities of which it is unnecessary to dilate 
here, further than to remark that its chief feature, the great 


development of the supra-orbital ridge, is sufficient to attract 
the attention of even the most untrained observer is commonly 
known as that of Neanderthal^ from a skull and some other human 
remains extracted in 1856 from a quaternary bed in the Feld- 
hofen cave of the Neander valley between Dusseldorf and Elber- 
feld, Rhenish Prussia. The cranium, pronounced by Huxley to 
be the most ape-like known (before 1892), is no longer regarded by 

K. 3 


Virchow as absolutely pathological 1 ; yet, although its normal 
character has been fully demonstrated by Broca, Brinton and others 
still clamour for its removal from anthropological works. But 
even so, they cannot get rid of the type, the general characters of 
which are shown in the restoration by Enrico Giglioli in his 
work on the antiquity of man*. 

In its essential features the same primitive type reappears in 
other well authenticated human remains, such as 
those of La Naulette near Dinant, Spy, also in 
Belgium, Shipka in the Balkan peninsula, Olmo in 
primitive Italy, Predmost near Prerau in Bohemia 3 , Sam- 

types, borombon and others in Argentina 4 and Brazil, 

lastly Trinil , in Java (Pithecanthropus erectus described at 

P- 144). 

From the American fossils, differing little from the forms still 

surviving in the same regions, Kollmann infers that the human 

species has not varied since quaternary times; whence his so- 

views f called Dauertypus, or "persistent type 5 . From the 

various an- European fossils similar inferences have been drawn 

thropoiogists. by Broca ^ de Quatrefages and the French anthro , 

pologists, who speak somewhat confidently of the persistence of 
the types of palaeolithic men, and of reversals to such forms 
amongst the present European populations. Broca even suggests 
that the French Kelts might have resulted from the fusion of 
quaternary man with Ligurian immigrants in neolithic times 6 ; 
while de Quatrefages affirms the survival amongst us of direct 
representatives of fossil races (Crania Ethnica, p. 28). So also 

1 Meeting of Berlin Gescll.f. Anthrof. Ethnol. &c., Oct. -20, 1894. 

2 I^UomOy sua antichitft &c., Florence, 1893, p. 9. 

8 The well-preserved fragments of skeletons of a whole diluvial family of 
six persons, found in 1894 by Herr Maschka, associated with the remains of 
numerous previously discovered mammoths; that of the man wonderfully com- 
plete and of gigantic proportions. (Athenaum^ Sept. i, 1894.) 

4 A skeleton with 13 dorsal vertebrae, as in Chimpanzee, Gorilla and Cebidre, 
instead of the normal is; found by Carles associated with a Megatherium 
(Vilanova, International Congress of Americanists^ 1892). 

5 Die Aittochtonen America* s % in Zeitschrift fur Et/inologie, 1883, and 

6 La race critique ancienne et moderns, in Rev. d*Anthrop. vol. I* 


Dr Houze', who considers that "heredity is so strong that, after 
thousands of years, we still find in the midst of our populations 
nearly pure descendants of our quaternary ancestors 1 ." 

At first sceptical, the foremost English anthropologists (the 
late Prof. Rolleston, Flower, Thurnarn, Barnard Davies, Galton 
and others) have now mostly accepted these conclusions, while 
still suspending their judgment regarding tertiary man. In the 
first report (1882) of the Committee appointed to obtain photo- 
graphs of the typical races in the British Isles, it is affirmed that, 
despite universal miscegenation, primitive racial types may still ' 
be recognised amongst the present inhabitants of Europe; that 
prehistoric characteristics do survive and that under favourable 
conditions a complete reversion to original types may take place 
through the operation of natural laws. The Committee seems 
to agree with Kollmann (the Swiss anthropologist) that original 
features may be detected even in mixed populations, owing to 
the persistence of pristine cranial characters long after the colour 
of hair and eyes had been blurred by crossing ; and lastly that a 
complete fusion of constituent elements never absolutely occurs. 
But there are reservations, and Dr Beddoe holds that, in the 
absence of information respecting the features generally, light 
hair and eyes and tall stature would not suffice to pronounce any 
person a Saxon, Dane or Swede. 

American specialists go even farther, and E. D. Cope, 
amongst others is inclined to regard the man of Spy "as a 
distinct species"; yet somewhat inconsistently thinks it equally 
probable that, "taking into consideration the characters of the 
Neolithic man, the Europeans originated in Europe, and that 
some of us are the direct descendants of the Homo neander- 
thalensis*" It should be added that this great palaeontologist 
does not derive man directly from the lemurs (as asserted by 
Topinard and others), without the intervention of the Anthrop- 
oidea. On the contrary his working hypothesis is that from the 
Eocene Lemur oids ascend the Anthropomorpha "which include 
the two families Hominidae and Simiidae" (ib. p. 326). In other 
words from a generalised Lemur type branch off the true Lemurs, 

1 Bull. d. I. Sac. d'Antkrop* de Bruxelles, 1894, p. 127. 

9 The Genealogy of Man in The American Naturalist, April, 1893, p. 335. 




the monkeys and the true Anthropoidea, which latter again 
ramify into the Simiidse and Hominidse as in our diagram 
(p. 19). 

Are we to conclude from all this that, unlike the types of 
other higher organisms, the human type and subtypes have 
persisted unchanged since early quaternary, if not late tertiary, 
times; in other words that man, highest and consequently most 
sensitive of all mammals, is not subject to the ordinary laws 
of physical evolution, that he remains unmodified under the 
thousand influences of the modifying environment, that his 
growth is out of harmony with the laws of the universe? The 
foremost expounders of evolutionary teachings would here seem 
to find themselves in direct antagonism to the fundamental 
principles of evolution itself, according to which not even species, 
much less varieties, are fixed and stable. Is there to be a 
Dauertypus of man alone, and perhaps of the molluscs near the 
other end of the gamut ? This is the first and one of the greatest 
difficulties that anthropology would appear to have passed on to 
ethnology for solution. 

The point may be understood, and perhaps explained, by 

reference to the somewhat analogous case of the 

tht E*qu!Itt. f Equidse, whose pedigree has been fairly well worked 

out from lower eocene times. Eohippus from the 

a. Pachynolophus* b. Anchtiherium. c. Anchitherium (late miocene). 
d. Hipparion. e. Equus (pleistocene). 

Mexican deposits of that epoch, a creature no larger than a fox, 


with four hoofed toes and an already atrophied thumb on the 
fore feet, is linked with the living forms (horse, ass, zebra, &c.) 
through Orohippus from the middle eocene of Utah and Wyoming, 
with four hoofed toes, Anchitherium from the early and late mio- 
cene with three hoofed toes in use, Hipparion from the pliocene 
with three hoofed toes, of which two are atrophied, and the 
pliocene horse with one hoof and traces of two others. Here 
we see the physical evolution of the Equidae completed within 
the tertiary epoch, since which time the further modifications 
have been confined to specific developments of their pliocene* 
precursor. Thus there has been continuous and orderly change 
from early tertiary to pleistocene or quaternary times, when the 
living species were established, all in strict accordance with 
evolutionary principles. 

So it would appear to have been with the Hominidse, except 
that here the postulated or actual (Castenedolo ?) ., _ 

* N . . ' Sergi s Tertt- 

precursor has developed quaternary varieties only, ary Hominidse 
and not true species. This is in accordance with rejcc e 
all the known facts, for Sergi's tertiary Hominidae are purely con- 
jectural and must be rejected, because their living representatives 
would necessarily be undoubted species, like the assumed tertiary 
forms themselves, and not merely varieties connected by un- 
broken intermediate links. Thus here also there has been con- 
tinuous change down to the quaternary varieties, and if these 
forms of the Hominidse are less specialised physically than the 
corresponding forms of the Equidae, the fact may be probably 
due partly to the more extensive intermingling of 

,5 - . ,. , Persistence 

the former, arresting further divergence, partly to O f, and Rever- 
the greater intelligence by which they were able 

better to protect themselves from the modifying 
influences of the environment. Thus may be explained Koll- 
mann's Dauertypus, the persistence of the human varieties already 
established in the quaternary epoch, and the reversals to the 
primitive types observed amongst the present European popu- 
lations, as now generally accepted by anthropologists. 

In the accompanying diagrams is shown the parallelism 
between the physical evolution of the Hominidae 

, ^ - , * , . . ,. Comparative 

and Equidae from their respective pliocene pre- diagrams of 



Pleistocene cursors. Here the greater divergence in the 
P liocene branches of the Equidae is seen to result 
in the present distinct pleistocene species, while 

the less divergent branches of the Hominidse indicate pleistocene 

varieties only : 

Pie istocene ] T arieti<.'* 

Pleistocene Species 



No difficulty is presented by the persistence of the human 
varieties since the pleistocene period, because, long though the 
interval be, it has still been too short for the development of 
specific differences even in the animal series. Consequently here 
no exception need be claimed, with Wallace for instance (Natural 
Selection, 1864 9) for man, whose physical evolution scarcely 
differs appreciably from that of his quaternary animal contem- 
poraries since the glacial epoch. Recent observations by such 
eminent palaeontologists as Albert Gaudry, Nehring and Boule, 
show that many living varieties of the cave fauna, the so-called 
Canis spel&us, Goldf. ; Lupus spel&us, Blainv. ; Hycena crocuta, some 
foxes, and even Leo spelaus differ scarcely more from each other 
(Mauritanian, Nubian, Persian and Gujerat lions, for instance) 
than they do from their pleistocene precursors; the divergence 
has consequently been no greater than has been that of the living 


human varieties from their pleistocene precursors 1 . Thus is 
reached the broad conclusion that the living Hominidae, pleisto- 
cene and pliocene man, with a postulated miocene precursor, 
would differ respectively as varieties, species, and genus, these 
organic differences themselves representing correspondingly longer 
geological epochs, in which to accomplish their several stages 
of development. 

1 Ce qui s'est produit pour 1'organisation de Tun s'est produit pour Forgani- 
sation de rautre...Le temps ecoule [depuis I'^poque quaternaire] a ete trop court 
pour que les accumulations de variations dans les organismes aient encore ptf 
operer des transformations de specificite', suivant la loi de la selection naturelle. 
Au contraire, les temps anterieurs se presentent a nous...comme ayant eu des 
durees incomparablement plus grandes, au cours desquelles la selection naturelle 
aurait eu le temps de manifester toute sa puissance " (K. Dupont, Bull, d. /. 
Soc. cPAnthrop. de Bruxdles, 1894, p. 328). 



Human incomparably greater than animal intelligence Growth of mind 
apparently out of proportion to that of its seat, the brain Evolution of 
organ and function correlated Cranial to be distinguished from mental 
capacity Comparative cranial studies often contradictory Chief physical 
determinants of mental power not so much the volume of the brain as 
its convolutions and the cellular structure of the grey cortex These 
elements capable of indefinite expansion till arrested by the closing of the 
cranial sutures Different degrees of intelligence in different races accounted 
for Such differences independent of the general bodily structure Hence 
physique and mental power not necessarily correlated and not always 
developed part passu But mind and cerebral structure always corre- 
spond Hence comparative study of brain texture, as by Broca and 
Miklukho-Maclay, yield best results Brain and its function, thought, 
capable of indefinite future expansion Differ in degree only, not in kind, 
from those of the lower orders Time alone needed to bridge the gap. 

IF the physical has been less, the mental evolution of man has 

been immeasurably greater, than in the Equidae, 

Human^in- the anthropoids and all other organisms 1 . To the 

greater than immense divergence in this respect is due the 

animal intelli- , - c L *. . ir* r\ i 

gence. tendency of former systematists (Cuvier, Owen and 

others), to separate man altogether from the other 

1 At the Exhibition of the Anthropological Sciences in Paris, 1889, a 
diagram was exhibited by Dr Topinard showing the great gap between the* 
brain weight and cranial capacity of man and the higher apes. Thus : 


Average for man 1*250) ,.. . 

b . [ grammes; difference, 0*874. 

Average for ape 0-376 j fe '* 

Difference between lowest human and highest ape species, o'6i2. 
Difference between lowest human individual (an Australian | 
woman) and highest individual ape J 

Capacity (volume). 

Average difference 885 c.c. ; difference between lowest human and highest 
Simian species 579 c.c, ; between lowest and highest individuals 377 c.c. 


mammals, and place him in a category apart. Indeed it might 
almost be said that for the ethnologist proper, who is concerned 
perhaps more with the mental than with the physical side, man- 
kind does occupy such a separate place in the animal king- 
dom. But the mind itself cannot be studied apart from the 
physiological conditions common to man and all other animals, 
else physiology becomes confused with psychology, and the 
subjective takes the places of the objective method, a retrograde 
movement opposed to the inductive philosophy as established by 
Francis Bacon, and foreshadowed by Friar Bacon, greatest of tht 
Schoolmen, who in times of pure speculation enthroned observa- 
tion, experiment, as "mistress of all the sciences, and end of all 
speculation" (domina scientiarum omnium et finis totius specula- 

Physiology has placed beyond all doubt the fact that the 
mental faculties are all localised in the brain; yet 
it might be supposed that there was no room in the Growth of 
skull for such expansion of its contents as would ^ y out^of 
seem to be required by the enormous progressive that o?lts seat 
expansion of the human intellect since the Homi- the brain. 
nidae branched off from the anthropoid stem. That 
there has been growth, and considerable growth, in size and 
weight has already been seen (p. 25) ; and in point of fact the 
lowest human brain stands in these respects as 3 to i compared 
with that of the highest apes. But such a ratio is a totally 
inadequate expression of the superiority of the human over the 
ape mind 1 . 

Have we here antagonism between function and organ, 
lending support to the view that the function 

creates the organ, and not the organ the function, " f 

in the idle discussion that has lately broken out on funct j on 

J correlated. 

this metaphysical point? Neither of these views 

1 The reader will observe that mind (thought, intelligence), not soul or 
spirit, or " psychic force " is here under discussion. Such expressions as 1 had 
a mind to do it; a strange thought passed through my brain, show how common 
speech shapes itself to the difference between these concepts, expressing the 
distinction between the ego and its functions more vividly than many psycho- 
logical essays. 


can be held, for at no stage of growth can one of the factors 
exist apart from the other. Function is an attribute, and 
attributes whiteness, hardness, reason, memory and the like 
can have no independent existence ; they are not entities but 
pure abstractions. Has anybody, except perhaps Plato, ever seen 
a tree that was not some kind of tree ? No, because tree is an 
abstract term which finds no place in the vocabulary of many 
primitive languages. If there was a hand (organ) it was used for 
prehensile purposes (function), because to grasp is an attribute of 
the hand, and by constantly exercising that attribute, that is to 
say, by dint of much grasping, aided by natural selection, heredity 
and time, the hand, starting as a rudiment, was ultimately 
perfected. Organ and function are developed together ; neither 
creates the other, at least in the physical world. Probably the 
prevailing confusion on this fundamental point of biological 
evolution, is to be traced to the prominence given by Lamarck to 
the "efforts of the inward sentiment 1 " of animals in bringing 
about changes in their physical constitution, thus suggesting the 
idea that "la fonction fait 1'organe" (M. Duval). 

So it is with brain and thought, and it might be inferred 

a priori that the human brain, seat of human thought, has been in 

some way physically perfected far beyond the ratio 

Mental c^- nd above indicated between it and the simian brain. 

pacity to be This inference is now demonstrated & posteriori by 

distinguished. i_-i-^ ui_ ^ *. j ^ 

physiologists, who show that mere size and weight, 
however important in themselves, are no measure of the real 
difference between the human and the highest simian brain. But 
owing to the relatively rare opportunities of studying the brain 
itself in its inner texture and structure, anthropologists have 
hitherto directed their attention mainly to the brain-case, deter- 
mining from its capacity that of the brain, and inferentially that of 

the mind itself. Thus the essential distinction 

Comparative ... . . 

cranial studies between cranial and mental capacity has been 

* l "Tout nouveau besoin...exige de 1'animal qui IVprouve, soit 1'emploi 
plus frequent de celle de ses parties dont auparavant il faisait moins d'usage... 
soit 1'emploi de nouvelles parties que les besoins font nattre insensiblement en 
lui par des efforts de son sentiment inte*rieur" (Philosophic Zoologique, I. p. 231 
of 1873 ed.). It is here that Lamarck and Darwin part company. 


confused, the most unsatisfactory and even con- often unsatis- 
tradictory results have been arrived at, the long factoi; y- 
and painstaking labours of many eminent comparative anatomists 
have been largely wasted, and craniology itself unduly discredited. 
" A few years ago it was thought that the study of Crania offered 
the only sure basis of a classification of man. Immense col- 
lections have been formed; they have been measured, described 
and figured; and now the opinion is beginning to gain ground 
that for this special purpose they are of very little value. Pro- 
fessor Huxley has boldly stated his views to this effect; and ih 
a proposed new classification of mankind has given scarcely any 
weight to characters derived from the cranium. It is certain, too, 
that though Cranioscopy has been assiduously studied for many 
years, it has produced no results at all comparable with the 
labours and research bestowed upon it. No approach to a theory 
of the excessive variations of the cranium has been put forth, and 
no intelligible classification of races has been founded upon it 1 ." 

The so-called cranial indices, on which a vast number 
of comparative measurements have been made between races and 
peoples, are concerned exclusively with the skull, the brain itself 
being seldom, and in certain cases such as fossil and sepulchral 
remains, never available. How valueless are many of the conclu- 
sions drawn from such comparisons may be inferred from the 
figures yielded by the skulls of Egyptian mummies, of neolithic 
man, and of Europeans at different epochs, often showing no 
change or even indicating apparent retrogression instead of pro- 
gress after long intervals of time. Thus Professor Schmidt* finds 
that the ancient Egyptian women had a higher cranial capacity or 
volume (1257 cubic centimetres) than the modern (1206 c.c.), 
while, according to Broca's measurements, the men of the 4th 
dynasty (1532 c.c.) stood higher than those of the i8th (1464 c.c.). 
This great anthropologist also determines the mean capacity both 
of neolithic and modern Europeans at about 1560 c.c. s ; no 
advance in some tens of thousands of years ! 

1 A. R. Wallace, Malay Archipelago, 5th ed. p. 599. 

2 Ueber alt- u. neuagyptische Schadel, in Archiv f. Anthrop. Brunswick, 
1887, vol. 17. 

8 Mf moires (CAnthrop. 1871, p. 334, and elsewhere. 


Yet the mental advance is unquestioned, and this advance- 

D rmi nient would be found to correspond with the physical 

nants of mental advancement, were the brain, and not merely the 

volume of S the C skull, of the early Egyptians and of neolithic man 

brain than its available for comparative purposes. But the con- 

convolutions * f r 

and cellular trast would have to be sought not so much in the 
structure. i / * i \ i 

respective volumes (size and weight), as in the 
sinuosities or convolutions of the inner white substance, and 
especially in the cellular structure of the thin outer cortex or 
envelope of grey matter which follows all the inner convolutions, 
with which it is also connected by an exceedingly complex 

nervous system. The countless cells of the cortical 
ments capable matter, fed through the nerves by the white sub- 
expanstoftm stance, are the ultimate seat of mental energy, and 
arrested by the here is found the field in which this delicate organ 

closing of the . ..... 

cranial su- and its function, thought, /may acquire indefinite 

turcs * expansion. The development of the cellular tissue, 

with a corresponding increase of mental power, apparently goes 

on until arrested by the closure of the cranial 

Different sutures. All the serratures are stated to be more 

degrees of in- ,-,,., , i 

teiiigencein complex in the higher than in the lower races, 

and their definite closing appears to be delayed till 
a later period in life amongst the former than 
amongst the latter. This physiological character, to which Filippo 
Manetta was the first to call attention in connection with racial 
differences 1 , has recently been noticed by two intelligent observers, 
Col. Ellis amongst the Upper Guinea peoples 2 , and Captain Binger 
amongst the West Sudanese populations generally. " The Black 
is a child," says this writer, "and will long remain so"; and the 
sudden arrest of the intellectual faculties at the age of puberty is 
attributed to the premature closing of the sutures 3 . Broca also 
has noticed that in idiots the " soldering " takes place early in 

1 La Kazza Negra nel suo state selvaggio <$rc., Turin, 1864, p. 20. 

2 The Ewe-speaking peoples > 1890, pp. 9-10. 

3 A cet arrt intellectual doit correspondre, dans ces regions [Sudan], la 
soudure de la bolte cervicale : le developpement du crane s'arrete et empeche 
le cerveau de se dilater davantage" (Du Niger au Golfe de Guinea 1892, 
II. p. 246). 


life, while the process is delayed the more the brain or mind is 
exercised. All this is highly significative, and would seem to place 
beyond doubt the direct relation between mental and cerebral 
expansion. An explanation is thus afforded of the different 
grades of intelligence that have everywhere been observed in the 
Hominidae throughout the historic period. In some the normal 
development of organ and function has proceeded simultaneously 
till the present highest level has been reached. In others both 
have been arrested at various stages by physiological causes, which 
may have had a pathological origin, but which more probably* 
represent various stages in the evolution of the cranium since 
quaternary times. This may be inferred from the fact that in the 
Neanderthal skull the frontal sutures are closed (ossified), while 
the occipital are more or less free, a distinct mark of inferiority, 
although the growth of the cerebellum is in no way determined 
by the premature closing of the frontal serratures. In the 
Marcilly-sur-Eure cranium, also distinctly palaeolithic, "la suture 
frontale a entitlement disparu, la soudure est complete; on ne 
voit plus de trace de dentelures sur le sommet de la teteV 
It would therefore seem probable, or at all events possible, that 
intense cerebration acts almost mechanically on the brain-cap, 
tending by its throbbing to keep the frontal sutures free till late 
in life, and even causing an expansion of the cranium itself in 
energetic and highly intellectual races. 

It is noteworthy that the mental differences are independent 
of the general bodily structure, so that the various 
groups of Homo Caucasicus, supreme in mental 

capacity, do not always compare with advantage pendent of the 

physically with some of the lower types; the South ture/ 
Europeans, for instance, with the Masai, the Zulu- Hence gene- 

Xosas, the Samoans and the Eastern Polynesians. ** 1 physique 

and mental 

Some of the highest intellects Alexander Pope, power not 
Heine and others have dwelt in feeble frames, "SSSdf 
while the stupid Serer Negroes of Senegambia are 
endowed with Herculean bodies. It seems obvious that since 
quaternary times the evolution of mind-stuff and of the general 

1 Philippe Salmon, Races Humaines prMstoriques, p. 15; -de Mortillet, 
LHomme, p. 48. 


physique has not gone on simultaneously everywhere, so that 
we now see great discrepancies and apparent contradictions 
between the two, one element progressing more rapidly in one 
place, the other in another. The explanation of such anomalies 
is perhaps to be sought in the very conditions of existence. 
In unfavourable environments, such as excessively hot and moist 
intertropical regions, man must either perish, or acquire such 
physical properties as may enable him to successfully struggle 
with the adverse climate, and in the struggle the animal is im- 
proved at the expense of the mental side. But in more favour- 
able surroundings, as in the temperate zone generally, the struggle 
is relaxed, the need of bodily perfection diminishes in direct ratio 
to mental growth, and the intellectual are, so to say, improved at 
the expense of the animal properties. But wher- 

an^ U ccrebr d i cver the P* nt can ^ e teste( i ^ w ^ probably be 
structure found that mental growth has always gone hand in 

respond? *" hand with an increasing number of cells of the 
cortex and of cerebral sinuosities. "That the 
convolutions in the Negro brain are less numerous and more 
massive than in the European appears certain 1 /* and it may be 
safely asserted that no brain of any inferior type will be found 
displaying the large number of windings exhibited by that of the 
late Professor von Helmholtz, one of the profoundest and most 
many-sided intellects of the age (ob. Sept. n, 1894). 

On the other hand the brain-cap of many savages has been 
found to be larger and heavier than that of some higher races. 
Thus in one of Broca's tables, quoted by Topinard (Anthropology, 
p. 230) the average of the neolithic skulls from the " Dead Man's 
Cave" is higher (1606 c.c.) than that of the Auvergnians (1598 c.c.), 
the Parisians (1558 c.c.) and the Spanish Basques (1574 c.c.), and 
not much less than that of Cuvier himself. After a long discussion 
of the subject in all its bearings, Waitz found himself " compelled 
to renounce the doctrine that the capacity of the cranium indicates 
the amount of mental endowment " (ib. p. 266). Nevertheless it 
will be seen that broad inductions may be drawn even from a 
comparison of cerebral volume alone. Years ago the subject 
engaged the attention of Dr C. G. Carus in connection with 

1 Waitz, Anthropology, I. p. 94. 


Morton's Crania Americatia, where the skulls of the Prairie 
Indians show a higher capacity (84 cub. in.) than those of the far 
more civilised Toltecs (77). If, says this deep thinker, \*e assume 
no direct relation between cranial volume and grades of culture, 
the former being merely or mainly the expression of a more 
vigorous bodily organisation, these discrepancies need not surprise 
us. It cannot be denied that general advance in culture is often 
attended by a certain decrease of physical strength (Abschwachung), 
by diminishing the need of bodily exercise, and removing us more 
and more from natural influences generally, thus inducing a 
certain delicacy (Verzartelung), and even sowing the seeds of 
many ailments by which the energies of the physical man are 
necessarily diminished 1 . 

Owing to the unsatisfactory results obtained from mere cranial 
comparisons, Broca himself turned later to the 

, , - .... , . , , . Henc - 

excllisive study of the brain, in which his greatest parative study 
triumphs were achieved 2 . During his residence turVpromUes 
at Brisbane, Miklukho-Maclay also began a syste- the best re- 
matic study of the cerebral structure from the 
ethnological standpoint. As was to be expected, he succeeded in 
determining a substantial difference between the brains of various 
races (Australians, Melanesians, Malays and other Mongoloid 
peoples) in the development of the corpus callositm, the pons 
Varolii and the cerebellum^ as well as in the relative development 
of the nerves, and in the grouping of the sinuosities in the great 
brain 3 . But the subject is in its infancy, although already suffi- 
ciently advanced to lead us to anticipate fruitful results from this 
new line of investigation. One point seems established, that as 
the sinuosities have had indefinite expansion in the . 

r . Brain and its 

past, so there is still ample room for indefinite function, 
expansion in the future. Consequently, although weoftade C ftStc 
size and weight may have probably attained their future expan- 
limits, the extent to which the cellular tissue may 

1 Enistehung und Gliederung der Menschheit^ 1858, p. 75. 

2 "Broca, il Pontefice Massimo dell' ipercraniologia moderna...non studia 
piu i crani, ma i cervelli " (Mantegazza, Archivio, X. 1880, p. 117). 

3 Nature, Dec. 21, 1882, p. 185. These results agree with those obtained 
by H. B. Rolleston from a careful study of the biain of an Australian, in 
yourn. Anthrop. Inst. 1888, pp. 32 et seq. 


improve is practically uncircumscribed, and in this direction a 
corresponding expansion of the mental faculties may be hoped 
for 1 . Ethnology properly understood affords no ground for the 
current pessimism, the disease of the age. By improved social 
institutions, more rapid progress may be made towards the ideal 
standard expressed by the formula, mens sana in corpore sano. 
Unless we take our stand on this firm ground of the simul- 
taneous upward growth of organ and function, 
itself the best guarantee of further upward growth, 

k thoseofthe we fall back either with Romanes on mysticism 

lower orders. - ** \ ^ *,- 

{Mental Evolution tn Man\ or with Mivart on 
supernatural intervention at the psychological moment (Origin 
of Human Reason). Separate brain from cerebration, and we are 
lost in the a priori reasonings of Noire and Max Miiller, ulti- 
mately rooted in the Kantian philosophy. The difficulties that 
all these evolutionists, and quasi-evolutionists, conjure up and leave 
unsolved, arise from the radical mistake of comparing the mind 
in its highest evolved state with that of the brute order, where 
the gap is so vast as to seem impassable without the extraneous 
aid of the supernatural or of metaphysics. They find in man, not 
merely sensation and receptivity, with perhaps a modicum of 
consciousness as in the brute, but true self-consciousness, which 
" enables a mind not only to know, but to know that it knows ; 
not only to receive knowledge, but also to conceive it... ; not only to 
state a truth, but also to state the truth as true" (Romanes, p. 
192). But in point of fact in the Fuegian, Tasmanian or Negrito 

mind there are the merest glimmerings of con- 

Time alone sciousness, and of self-consciousness next to nothing. 

bridge the gap. The Fuegian "self-consciousness," for instance, may 

be gauged by the Fuegian " conscience," which in 
stormy weather flings wife and children overboard, to lighten the 
overladen craft of so much freight, not, as has been said, to 
propitiate gods (or demons) of whom it knows nothing 8 . These 

1 Topinard, UHomme dans la Nature. Chap. XXII. 

2 Lovisato, who has made a careful and most sympathetic study of these 
aborigines, denies them retentive memory, and compares their intelligence to 
the stationary instincts of animals: "I Fueghini hanno poca intelligenza, 
pochissima memoria, nessuna ritentiva. La loro abilita [mental capacity] puo 
essere per alcuni rispetti comparata agli istinti degli animali, perche non e 


are the minds, and not those of Plato, Kant, Newton, Darwin, 
that should be compared with those of man's nearest congeners > 
then it would be seen that the difference is one, not of kind but 
of degree only, a difference quite capable of being bridged over 
by natural process of development in the course of a few hundred 
thousand years. It is time, not metaphysic or miracle 1 , that is 
needed. In the next chapter an attempt will be made to provide 
this time, the great instrument with which Nature works out all 
her transformations. 

migliorata dall' esperienza...Difficilissimo e per loro comprendere le piu sem- 
plici alternative, quindi con immensa difficolta si possono da loro avere delle 
informazioni, e non si puo mai essere sicuri se, a furia di domande, essi abbiano 
esattamente compreso cio die loro abbiamo detto " (ib. p. 27). Yet this group 
(the Yahgans) had at that time (1870 1884) been fourteen years in the hands 
of the English missionaries, and nearly all were " Christians." Since Darwin's 
time ( Voyage of the Beagle) there had been no appreciable intellectual advance- 
ment, despite the flourishing reports received in Europe, and despite the 30,000 
words with which their language was strangely credited. 

1 Even Joseph Le Conte, whose theory of evolution is a ' * Christian pan- 
theism " worked out by " paroxysms," allows the domain of science "to remove 
as much as possible the miraculous from the realm of nature " (Man's Place in 
Nature, Princeton Review, 1888, p. 784). This is considerate; only "the 
realm of nature," like Oliver Twist, asks for more. In fact evolution is a 
jealous mistress; she will have all or nothing. 



The Geological Sequence in its bearing on the Antiquity of Man Table of the 
Geological Sequence: Primary; Secondary; Tertiary; Quaternary; Pre- 
historic; Historic The Glacial Problem Reactionary Views Croll's 
Periodicity Theory Objections and Limitations of Time by Prestwich A 
reductio ad absurdum Arguments based on influence of Gulf Stream and 
Absence of Glaciation in earlier geological epochs estimated Croll's Theory 
reaffirmed A long period of time needed to meet all the conditions : Re- 
distribution of Land and Water; Intermingling of Arctic and Tropical 
faunas; Scouring out of great river valleys; Man long associated with 
extinct animals; Britain twice submerged since its occupation by man; 
Little trace of primitive man in the last post-glacial deposits of the North 
Two Ice-ages and long Inter-glacial period essential factors Difficulties 
of the Intermingled Arctic and Tropical Faunas Lyell and Boyd-Dawkins' 
"Seasonal- Migration" Theory discussed Long association of reindeer 
and hyaena explained Great age of the flints found in the high-level 
drift, boulder-clay, plateaux and riverside terraces Pre-, Inter- and Post- 
Glacial Man The problem restated General Conclusion Pliocene 
Hominidse rejected Specialised Inter-glacial Hominidae reaffirmed 
Their probable age Post-glacial Man a nondescript. 

IN the foregoing chapters the age of man was rather anticipated 
than determined. Before discussing the subject 
caTsequence^n m re fully, it will be convenient to define the some- 
tife b uestion < of w ^ at P uzz h" n g geological terminology bearing on the 
the antiquity of question of man's first appearance on the globe, so 
man * that the non-geological student may have some 

idea of what is meant by such expressions as tertiary or quater- 
nary man, pliocene or pleistocene times, and so forth. This will 
best be done by giving a table of the geological sequence as 
recorded by the succession of stratified rocks in the crust of the 
earth, tabulating more fully the later periods with which we are 
here more immediately concerned. Geologists and solar physicists 
are not of accord as to the probable duration of the whole process. 


Nothing can be definitely known ; but it would seem that if the 
geologists a'sk too much, say, a round 100 m. (m = million) of years 
the physicists grant too little (20 m. or even less). Most, however, 
agree that the earlier stages took far more time in running their 
course than the later, the decrease of time proceeding by a sort 
of geometrical ratio down to the contemporary period. Hence if 
50 be taken as a rough compromise between the extremes, the 
duration of the several epochs in the evolution of the earth's crust 
will be approximately as here indicated 1 . The point has an 
obvious bearing on the question of man's antiquity, as far as it 
can be measured in terms of years. The ratios of geological time 
are based on Sir A. Ramsay's estimate, which assigns 79 per 
cent, to the palaeozoic system, 18 to the mesozoic, and 3 to the 
cainozoic, these being respectively about 57,000, 13,000, and 
2,240 feet thick 2 : 


i. Laurentian and IJuronian systems, largely developed in 
the Lower and Upper (Lake Huron) St Lawrence 
Basin, whence their names ; over 30,000 (?) feet geological 

thick : reddish gneiss ; stratified crystalline rocks, 



mica schists, quartzites; unfossiliferous limestones 

1 Being mere ratios, the figures here given can of course adapt themselves 
to any view ; those claiming 100 m. need but double them, while those satisfied 
with less can halve or quarter them to fancy, always, however, bearing in mind 
not only the slow growth, but also the subsequent weathering and disintegration 
of vast geological formations. In the Sahara alone some two million square 
miles of unstratified and sedimentary rocks have thus been triturated probably 
since secondary times. They were not deposited, as generally supposed, on a 
marine bed, and then upheaved. The sands of the Desert contain no marine 
fossils (Suess, Reclus, Playfair). 

2 jfotirn. Geological Soc. 1860. No doubt these proportions have recently 
been questioned, and the extent of the palaeozoic system especially is believed 
to have been exaggerated. But until replaced by something definite, they may 
here be retained, all the more that the question of man's antiquity is not affected 
by the greater or less duration of the early geological eras. It may be added 
that Prof. John Perry seems now to have satisfied Lord Kelvin that his original 
"20 millions" are far too little as the shortest limit: "I should be exceedingly 
frightened to meet him [Geikie] now with only 20 millions in my mouth" 
(Letter to Mr Perry, Nature, Jan. 3, 1895, p. 227). 

- 42 


and plumbago, of doubtful organic origin ; but Eozoon canadense, 
long supposed to be an organism (rhizopod) is now believed to be 
a mineral structure ; Huronian is intermediate between Laurentian 
and Cambrian ; duration 20 m. 

2. Cambrian, from Cambria = Wales, where first studied : 
Llanberis, and other slates, schists, sandstones and conglomerates ; 
fossiliferous, containing numerous invertebrate organisms, such as 
trilobites and brachiopods ; duration 8 m. 

3. Silurian, from the pre-Aryan peoples of parts of Britain 
known to the ancients as Silures: Llandeilo flags and schists; 
Caradoc sandstone ; Dudley (Wenlock) limestone ; Ludlow lime- 
stone and shale ; Brecon and other tilestones ; in Germany some 
greywackes (Gramvacke); in America the Trenton, Niagara and 
other formations, certainly of Silurian age ; in Russia widely 
diffused; brachiopods, trilobites and other Crustacea; algae, corals ; 
also fishes, but no higher vertebrates ; duration 5 m. 

4. Devonian, named from the characteristic Sandstones and 
limestones (marine) of Devonshire, perhaps later than the Old Red 
Sandstone of Scotland : probably some beds are lacustrine (S. Wales, 
Shropshire, &c.); all nearly synchronous ; analogous sandstones in 
Russia, France, Belgium, United States, &c., with similar fossils, such 
as corals, brachiopods, and especially Ganoid fishes; duration 3^ m. 

5. Carboniferous : carboniferous limestones, millstone grit, coal 
measures, calciferous and other sandstones, deposited in lagoons, 
estuaries and surrounding seas ; fossils abundant ; corals, molluscs 
of all the known orders, spiders and certain insects ; many fishes ; 
labyrinthodon and other amphibia ; duration 2 J m. 

6. Permian, named from the characteristic rocks of Perm 
(Russia) : variegated sandstones, magnesian limestones, marl slate, 
red sandstones and clays ; rich floras and faunas ; numerous 
genera of fishes and reptiles; but no true birds or mammals; 
duration 2 \ m. v 


7. Triassic, from Gr. trias three, in reference to the three 

series determined in Germany, in ascending order : 

Secondary. ._ , ._, . _ __ " 

Bunter sandstone (Gres des Vosges of France, 
mottled, red and other sandstones in England); Muschelkalk, a 


marine limestone, wanting in England; Keuper sandstone (new 
red marls, &c.); molluscs, amphibians, reptiles; a few mammals; 
duration 2 m. 

8. Jurassic, from the Jura mountains; corresponds to the 
Liassic and Oolitic of Britain : Fuller's earth of Bath ; Oxford 
clay; Kelloway rock; Portland and Purbeck beds; blue and 
grey limestones ; most liassic fossils marine ; enaliosaurians, 
ichthyosaurians, plesiosaurians, numerous fishes; oolitic fossils, 
reptiles, marsupial mammals, one bird (archaeopteryx) ; Britain 
and West Europe largely under water; duration i \ m. 

9. Cretaceous, "chalk-like": Wealden, Neocomian (Green- 
sands); gault; white chalk; Maestricht beds; ammonites; igua- 
nodon; pterodactyls; probably several true birds. Throughout 
secondary times India was probably connected by continuous land 
across the Indian Ocean with South Africa ; duration if m. 


10. Eocene, "dawn of new" forms, in reference to molluscs: 
(a) Loivtr Eocene: London clay, Paris gypsum, 

Barton sand and clays; (b) Oligocene or Upper 
Eocene, represented in the Hampshire basin, largely developed 
in Germany (Maintz beds, &c.); all still surviving invertebrate 
classes ; crocodiles, tortoises, a few birds ; the living orders and 
families of mammals (Palseotherium, Anoplotherium, lemurs, &c.); 
duration 1*25 m. 

IT. Miocene, "less new" forms (of molluscs): (a) Lower 
Miocene, equivalent to the Upper Oligocene of the mainland ; (b) 
and (c) Mid and Upper Miocene; Thenay deposits; marine 
miocene of the Mediterranean, Egypt, India and Australia; 
Western parts of the United States; many higher mammals, 
including many living genera: Dinotherium giganteum, Mastodon 
angustidens, Rhinoceros Schleiermacheri, Machairodus (sabre- 
toothed tiger); Pliopithecus, allied to Gibbon; Dryopithecus, 
apparently allied to Chimpanzee ; remote precursor of man postu- 
lated; in America Mesohippus, Miohippus, Perchoerus, Elotherium, 
Hyaenodon; duration i m. 

1 2. Pliocene, " more new " forms : two divisions ; (a) Old 
Pliocene: Red Crag pf Suffolk ; white (Coralline) Crags ; Antwerp 


Crag; Pikermi beds near Athens; Deposits of Sivalik (Hima- 
layan foothills); N. American Pliocene, (b) New (Late) Pliocene: 
Norwich Crag; Forest beds of Norfolk cliffs; German, French, 
Italian pliocenes; many genera of mammalia, and some living 
species, including all the anthropoid apes, man's immediate pre- 
cursor, and man himself; Elephas antiquus, E. primigenius, 
Rhinoceros tichorhinus (Woolly r. of the Thames Valley brick- 
earths) ; R. leptorhinus ; Machairodus ; Hippopotamus ; Cave 
lion ; Cave bear ; spotted and striped hyaena ; Irish elk ; Bison 
priscus ; all these widespread throughout Europe and Britain, 
some surviving into next period; Britain connected with main- 
land through the East Anglian forest beds, which extended along 
Norfolk and Suffolk coasts, passing under the present cliffs and 
beneath the North Sea to the Continent, the land being also 
continuous westward through Ireland to the 100 fathom line in 
the Atlantic; climate warm or mild till towards the close, when 
first glacial epoch sets in; duration 0-850 m. 


13. Pleistocene (" most new "), a term used somewhat vaguely 
by English and foreign geologists, the line being 
p?eistocenej[ difficult to draw between this and the new pliocene; 
distribution of land and water nearly the same as 
it present; but northern hemisphere subjected to two or more 
nvasions of ice (several according to Prof. Geikie), with inter- 
vening warm epochs of varying duration. Formerly these glacial 
Deriods were supposed to be synchronous with subsidences, by 
vhich Britain was once if not twice severed from the mainland ; 
jut many geologists now believe that the Ice Age was more pro- 
bably coincident with elevation rather than subsidence, and in 
my case it is extremely doubtful whether the connection of Britain 
vith the Continent was interrupted in early pleistocene times. 
During the recurrent ice invasions, which caused the extinction of 
many pliocene forms, there was a perplexing association of 
tropical, temperate and apparently arctic animals, the more 
:haracteristic pleistocene fauna being : Hominidae, already 
spread over most of the dry land throughout the whole world 


(PALAEOLITHIC MAN); lion, brown and grizzly bear, hyaena, rein- 
deer, panther, Kafir cat, lemmings, varying hare, musk sheep, 
glutton, arctic fox, alpine snowy vole, chamois and ibex; mam- 
moth, urus. Chief formations : brick-earth, fluviatile loam, high 
level gravels, loess; cavern and glacial drift deposits 40 to 
60 feet thick, variously known as diluvium, boulder formation, 
boulder-clay, boulder-drift, or simply drift; all unstratified detritus 
(clays, flints, gravels, sands, shingle), transported partly by ice- 
bergs, partly by land-ice, and deposited during both glacial epochs. 
Cave dwellings, kitchen-middens, stone workshops, rude chipped 
and unpolished stone implements of simple types (spear-heads, 
scrapers, hammers, &c.), very much alike everywhere, chiefly of 
flint, chert, quartzite and ironstone, besides awls, borers and 
other objects of bone or horn, showing in some places distinct 
progress in the arts of palaeolithic man ; approximate beginning of 
strictly pleistocene or quaternary times 600,000 or 700,000 years 
ago ; duration 0^530 m. 

14. Post-Pleistocene, or Prehistoric: measured by the raised 
beaches, peat-bogs, alluvial deposits in Mississippi p os t-pieis- 
basin, Nile delta, and elsewhere and by other con- tocene (Pre- 

... . historic). 

siderations, can scarcely be less than 60,000 years, 
probably more ; largely coincides with the general disappearance 
of ice and the appearance of the Men of the New Stone Age 
(NEOLITHIC MAN), supposed to be separated in some places by 
a considerable interval from their palaeolithic precursors, but 
both more probably merged together by insensible transitions; 
domestic animals, cultivated fruits, primitive arts (agriculture, 
pottery, spinning, weaving, mining of copper, tin, gold), shell- 
mounds, lake dwellings, barrows, sepulchral chambers, rude 
megalithic and monolithic monuments (menhirs, dolmens, crom- 
lechs), some types stretching nearly round the globe from the 
Naga and Khasi Hills through North Africa to West Europe and 
Britain, and across Atlantic to Tiahuanaco (Lake Titicaca); 
camps, fortified earthworks, polished and perfected stone im- 
plements, dolabra, celts of varied forms and use, chiefly of flint, 
chert, dolerite, quartzite, obsidian, jade, jadeite or nephrite; in 
the eastern hemisphere neolithic man passes successively and 
without interruption ind^Tthe copper, bronze and iron ages, repre- 


sented in the New World by copper only, these ages merging 
generally in the north temperate zone in the 

15. HISTORIC (CONTEMPORARY) PERIOD: chronology, written 

records, literature, fine arts, modern culture con- 
Historic* , , . . , . ~ . 
nected by vague reminiscences and some definite 

traditions with the bronze and earlier ages 1 . Duration of historic 
period (in the Nile Valley and Mesopotamia) scarcely less than 
10,000 years. With the progress of archseological research the 
beginnings of Egyptian culture recede farther and farther into the 
remote past. Menes, reputed founder of Memphis and of the 
first empire (5000 B.C.; Mariette), was recent compared with the 
builders of the rude monuments brought to light by Mr Flinders 
Petrie at Coptos, Upper Egypt, in 1894. Mr Norman Lockyer 
also 2 shows on astronomic grounds that a temple at Edfu towards 
the Nubian frontier was built for the observation of the star 
Canopus about 6400 B.C.; a date requiring the beginnings of 
Egyptian culture to be extended much farther back than had 
previously been supposed necessary. 

Before attempting to determine the date (geological) of man's 
advent, it will be necessary first to deal with the 
1 Glacial problem, around which the battle has raged, 
and still rages, especially in Britain and those other 
northern lands where naturalists and archaeologists are brought 
into more direct contact with glacial phenomena. 

Lately there has been a reaction, led by Prof. Prestwich, 
against the conclusions which appeared to have been firmly 
established and even virtually accepted by Prestwich himself 3 , 

1 Homer and Hesiod, earliest names in Greek literature, of unknown date, 
and therefore in a sense prehistoric, had memory of times when bronze only 
was in use, iron being still unknown : 

rots 5* rp x<fcX*ea ^v retfxi, x^ KfOL 
Xo>\K(f d j cipydfrvro, ^Xas 8' oifK tone 

Works and Days. 

* The Dawn of Astronomy, &c., 1894. From the section devoted to the 
early Babylonians it would appear that their independent astronomic obser- 
vations were not less ancient than those of the Egyptians. 

* "I am disposed to consider with Mr Tiddeman, that the cave which he 
is now investigating at Settle [the *' Victoria" Cave, Yorkshire] may be of pre- 


regarding the considerable antiquity of the ice age and the exist- 
ence of pre-glacial man. From the above table of geological 
sequence it would appear that the glacial periods, beginning in 
late pliocene times, reached far into the quaternary, and including 
the interglacial interval, may have lasted altogether some 300,000 
or 400,000 years. Adding the remainder of the quaternary and 
the whole of the contemporary, this would give an antiquity of at 
least half a million years to pre-glacial man, if he existed ; at least, 
because the time he may have existed before the first appearance 
of the ice would have still to be considered, although necessarily 
undeterminable even approximately. There are no data by which 
it could be limited except the first appearance of the allied forms 
(Pliopithecus, Dryopithecus), which go back to the miocene, when 
a remote and generalized precursor of Hominidae must indeed be 
postulated 1 . But the character of such a precursor, whether truly 
human, or only a homo alalus, would remain a subject of pure 
speculation, interesting to systematists, but of no importance to 
the ethnologist, who must limit his inquiries to pre-glacial man, 
as above roughly determined. 

When the problem of man's antiquity is put in this way, clear 
of all side-issues, the importance is at once seen of 
further determining the approximate first appearance dicitytheory~ 
and duration of the ice age. This problem was first 
seriously attacked by Dr Croll in a classical work 2 , the conclusions 
of which have been questioned chiefly by those who are strangely 

glacial age." (Journ. Anthrop. Inst. 1877, p. 177.) By pre-glacial, however, this 
geologist does not mean a separate period of indefinite duration preceding the 
glacial, but only the earlier stages of the glacial itself, which he divides into 
three periods : pre-glacial, that of first increase ; glacial, that of maximum cold ; 
post-glacial, that of last decrease (Jotirn* Geolog. See. August 1887, p. 404). 

1 Some philosophers object to anything being "postulated," except perhaps 
where miracles may be needed to bridge over gaps. But if we see one section 
of a chain suspended from above, and another advancing to meet it from below, 
and if the point of junction happen to be hid from view by some intervening 
obstacle, we all unhesitatingly concede the invisible third section; we "postu- 
late** the "missing link." So in the orderly succession of organisms in the 
realm of nature, where the missing links, which have excited so much senseless 
ridicule, must also be granted; else the lower series has no end, the upper no 
beginning, but hangs dangling in nubibus, the greatest miracle of all. 

2 Climate and Time, 1875. 


reluctant to concede a long term of existence to man, the noblest 
outcome of organic evolution. Reasoning somewhat after the 
manner of Kepler, who by pure mathematical computation deter- 
mined the afterwards verified interplanetary distances, Croll, be- 
lieving that periodical changes in the eccentricity of the earth's orbit 
afforded the best clue to great secular variations of climate, and 
that the periods of greatest cold coincided with those of greatest 
eccentricity, determined the maximum and minimum periods for 
3,000,000 years backwards and 1,000,000 forwards from the 
present time. He thus found that within the last million years, 
say, since near the close of the pliocene, there had been two 
maxima, one lasting 260,000 years (from about 980,000 to 
720,000), the other 160,000 (from about 240,000 to 80,000 years 
ago), the differences in duration being due to differences in the 
maxima themselves. At first Croll with " several eminent 
geologists" referred the glacial epoch proper to the earlier and 
longer period ; but afterwards, in order to be well within the 
limits, and also perhaps to conciliate prejudice, he made the 
glacial epoch coincident with the later and shorter period, closing 
about 80,000 years ago, and giving to pre-glacial man an age of 
not necessarily more than about 240,000 years (the beginning of 
the last maximum of eccentricity). Then much ingenuity was 
exercised in the effort to reconcile this limited period with the 
great changes that have since taken place, especially in the 
physiographic distribution of land and water. 

But now Prof. Prestwich will not even grant so much, and 

whittles down the last glacial epoch which he has 
and limitations never worked out mathematically, to "from 15,000 
Prestwich to 2 5' 000 years," and the post-glacial "to within 

from 8,000 to 10,000 "j and he adds: "This might 
give to Palaeolithic man... no greater antiquity than perhaps about 
from 20,000 to 30,000 years; while should he be restricted to 
the so-called post-glacial period, his antiquity need not go 
farther back than from 10,000 to 15,000 years, before the time of 
Neolithic Man 1 ." And thus we get within measurable distance 
of the Mosaic Cosmogony. But such a reductio ad absurdum is 
not reached without some straining and even distortions of 

1 Joum. Geolog. Soc. August 1887, p. 407. 


physical facts, as, for instance, when the Gulf Stream is in- 
troduced to help in getting rid of the great ice-sheet some 
thousands of feet thick, and made to flow in larger volume 
than at present through the then wider Florida Strait (ib. p. 403). 
But where did the larger volume come from? Were Florida 
Strait now again widened, the volume would not be increased by 
a single drop, but only spread over a correspondingly wider space. 
Or if the narrowing of the channel diminished the outflow, as 
implied, then the waters that failed to escape would be dammed 
up in the Gulf of Mexico, upsetting the general equilibrium of th% 
ocean level, were such possible, and developing in Florida Strait 
a prodigious series of falls and rapids on a scale large enough to 
dwarf ten thousand Niagaras. But in point of fact only a very 
small portion of the Gulf Stream penetrates through the Lesser 
Antilles into the Caribbean Sea, and so round the Gulf of Mexico 
and through Florida Strait to the Atlantic, where it merges in 
the much larger body which never enters the narrow seas at all, 
but skirts the north side of the Greater Antilles westwards to the 
confluence. Any local changes could not consequently affect 
its volume, and inferentially the climate of the higher latitudes. 

But Prof. Prestwich goes still further, and argues against 
Croll's periodicity theory of the earth's eccentricity, on the ground 
that very few or no traces of glaciation can be detected in 
earlier geological epochs, the Chalk, for instance, the Carbo- 
niferous, "the Devonian and Silurian periods," say, as far back as 
15 or 20 million years ago 1 . But there seems here confusion of 
cause and effect, a fruitful source of error in all such reasonings. 
Croll carried his calculations only 3,000,000 years back, and had 
he gone further, he would, or at least should, have warned his 
opponents not to look for much results when ages of exceeding 
high temperature were reached. The fierce cyclone, which cuts 
like a knife through the Alleghany woodlands, strews the prairie 
with the wreckage of mushroom towns, and churns up the waters 
of the Great Lakes, sweeps almost harmlessly over the stoutly 
built cities of the Northern States. So with the recurrent periods 
of cold, which in recent and temperate times covered vast areas 

1 Or a proportionate increase or reduction according to the various views 
held regarding the absolute duration of geological time (see note, p. 51). 


with great ice-caps, but which must have had continuously 
diminishing effects the farther back their action is looked for. 
They may no doubt have stimulated the cooling process; but 
they could scarcely have had any appreciable effect, for instance, 
in pre- Archaean times, when the surface of the earth may be 
supposed to have still been in a somewhat viscous or semi-fluid 
state, owing to the great internal heat thousands of aeons ago. 
Nor could great eccentricity have laid down thick-ribbed ice even 
in the early Archaean ages, when the crust, though hardening, was 
Still too warm to support any but the lowest organisms. Thus 
we see that maximum eccentricity is merely a determining cause, 
whose varying effects will depend on the varying conditions. 

But whether the causes be cosmic or telluric, or, as would seem 

Croii's self-evident, both, complementing and reacting on 

theory re- each other, Croll's last two glacial epochs must be 

rmc " accepted in all their fulness and not attenuated 

down to the level of narrow preconceived views. They are 

needed to account for a thousand other facts in the natural 

history of the Hominidae, which are apt to be overlooked when 

this exceedingly wide field of research is surveyed with reversed 


It must however be confessed that it is no light matter to keep 

steadily in view a vast horizon, which in space coin- 

of time*ieded cides with the terrestrial periphery, in time stretches 

to meet ail the k ac fc f or Sav h a jf a m inion years ; a horizon which 
conditions. * ' 

covers two, if not more, glacial epochs with an inter- 
vening period of some hundred thousand years. And can less be 
demanded, is it possible to move freely in narrower limits, where 
we have to consider the overlapping of the geological frontiers 
(pliocene holding its ground long after pleistocene has dawned); 
great disturbances of the planetary surface, such as the Indo- 
African breaking up and reappearing as the Indo-Asian con- 
tinent ; possibly high land rising and again vanishing between 
Europe and the New World, and Britain presenting great oscil- 
lations of level before its final severance from the mainland; 
apparent intermingling in Britain itself and on the Continent 
of reputed arctic and decidedly tropical faunas, such as shaggy- 
maned mammoth, large and pigmy hippopotamus, sabre-toothed 


tiger, huge cave bear, superbly antlered Irish deer, some dis- 
appearing, others persisting till joined by fresh arrivals such as 
the lion, spotted hyaena, bison, musk sheep, and reindeer ; lastly 
the scouring out of great river valleys, such as Nile, Somme, 
Thames and Rio Negro, down to depths of over a hundred feet, 
and even fourteen hundred (Nile), to their present levels, while their 
banks were frequented and their caves inhabited by palaeolithic 
hunters ? For man had already arrived, was witness to many of 
these shifting scenes, looked out on arctic seas strewn with icebergs, 
fashioned some of his rude flint and quartzite weapons from th* 
very materials borne down by those floating masses, took refuge 
with those strange beasts in the caves of Devonshire, Derby- 
shire, France, Belgium, Italy and Germany, and, after the retreat 
of the first ice-stream, associated with them on the sunny plains 
of inter-glacial times long enough to improve his processes and 
enable us to distinguish grades of culture even in the palaeolithic 
period, long before the appearance of Neolithic Man on the 

And all this before the last and shorter invasion of ice, which 
again drove man and beast for shelter to their old cavernous 
haunts. Else how explain the orderly succession of fossil and 
tool-bearing deposits in those retreats, always ascending from 
rude beginnings to more perfected workmanship 1 , and even to 
skilful carvings of human and animal forms, on the bones of 
the extinct or living species themselves, as seen for instance in the 
specimen from the Oeswell cave, Derbyshire (well-designed head 
of a "hog-maned" horse on a rib-bone), and in greater variety and 
higher artistic skill in the Dordogne Caves, France, still before 
Neolithic times ? And we are asked by professors of geology, who 
understand better than the lay mind what all this means, to 
believe that possibly the age of palaeolithic man " need not go 
farther back than from 10,000 to 15,000 years before the time oi 

1 The rudeness of the implement, however, is not always a proof of greater 
age ; the neater finish may sometimes be due to a more easily worked material, 
so that a rough quartzite so hard to manipulate, may be more recent than a fine 
flint object. " In some part of the older deposits of St Acheul the flint imple- 
ments are better made than the newer ones found in the lower gravels of the 
Somme Valley " (Prestwich). But where all are of one or the other material, 
as is often the case, no doubt arises. 


Neolithic Man." If this be so, then Neolithic Man himself must 
be put back about half-a-million years or so beyond his probable 
advent, and for this there is no kind of evidence. 

Appeal is made to a possible, and no doubt probable, more 

copious rainfall formerly than at present, by which 

subme^e'd* tlie scourm g process might be accelerated. But 

since its rainfall must have its limits, say, at the outside 500 

occupation by s i 11 i r^ 

man . or 600 inches annually, as at present on the Garo 

and Khasi Hills, Brahmaputra basin ; else neither 
nan nor beast could exist in the Thames or Somme Valleys. 
But what effect could any amount of rainfall have, for instance, 
on the upheaval and subsidence of the land itself, rivers and all, 
since the arrival of palaeolithic man in Britain? For we are in 
the presence of such phenomena, all admitting that primitive 
man roamed South Britain, if not synchronously with the gradual 
subsidence of the late pliocene East Anglian fir, spruce, oak and 
birch forest beds beneath the German Ocean, at least during the 
formation of the old Thames deposits in Kent and Essex. Here 
man lived in association with such fauna as the lion, spotted 
hyaena and musk sheep, as proved by the flints discovered by the 
Rev. Osmund Fisher at Crayford, by Messrs Cheadle and Wood- 
ward at Erith, and still more by the very workshops where these 
implements were made, as revealed by the researches of Mr Flax- 
man Spurrell. 

This must have been in early pleistocene times, before the 
land-ice had advanced southwards, strewing a great 

e m f part f the SUrface with itS boulder-clays. As 

the last post- pointed out by Mr W. Shone, the early hunters de- 
pots. C " scribed by Boyd Dawkins as contemporaries of 
"the leptorhine rhinoceros, hippopotamus and 
straight-tusked elephant," were certainly pre-glacial, that is, ante- 
cedent to the last ice age. "If Britain were inhabited by this 
early race of palaeolithic hunters they must have traversed large 
areas of the surrounding country in search of food. In the 
excitement of the chase it is inconceivable that they should not 
have left many a lost weapon, which would to-day testify to their 
existence in post-glacial times. Taking England from the Mid- 
lands to Berwick-on-Tweed was ever country so delved to make 


roads, harbours, mines, canals, railways, and great towns and 
cities ; yet no trace of this palaeolithic hunter or of his contem- 
poraries 'the leptorhine rhinoceros, hippopotamus and straight- 
tusked elephant ' has ever been found over this area upon a post- 
glacial surface or in post-glacial strata 1 ." 

Boyd-Dawkins argues that because the raw material of some of 
the implements appears to have been obtained from the glacial de- 
posits, early man must have made his appearance " after the district 
[Pont Newydd near St Asaph] was forsaken by the glaciers and 
the sea 2 ." But, as shown by Mr Morton in this and obviously in 
many analogous cases, the stone may have equally 
well been obtained "from its original locality, or from Two ^ e 

J 7 ages and warm 

moraines in the early period before the district was inter-glacial 
submerged, as it had been brought near to the cave Sa" factors."" 
in the form of boulders 3 ." It is obvious that the 
broad question of Pre- 9 Inter- and Post-Glacial Man cannot be 
intelligently argued on such narrow issues as these. But such dis- 
cussions, which but increase the perplexity of the subject, will 
doubtless continue to prevail until it is raised to a higher plane of 
thought by the frank acceptance or denial of Croll's two ice ages 
with a long intervening warm inter-glacial period, on which every- 
thing depends. The point is put very clearly in her treatise on the 
Paris Basin by Mme. Cldmence Royer, who shows that glacial 
phenomena were of two kinds, one characterised by the presence 
of floating ice (icebergs) and a considerable subsidence of the 
land, the other marked by great altitudes and glaciers descending 
far lower than at present. The chief polar phenomena (first age) 
occurred " between the miocene and the pleistocene 4 ." How 
could the Swiss glaciers, for instance, descend much lower than at 
present during the first age, when the plains were under water ? 
Hence their greater former development, admitted by all, must 
have occurred at some other time, in fact during the second ice 
age, when the land stood at a much higher level above the sea. 

1 Post- Glacial Man in Britain, in Geol. Mag. 1894, p. 79. 

2 Early Man in Britain, p. 192. Elsewhere of course this authority admits 
that the River-drift hunter may have been "quite as likely pre- as post-glacial" 
(On the Present Phase of the Antiquity of Man, Meeting of Brit. Ass. 1882). 

3 Geology of the Country around Liverpool, quoted by Shone, to. 

* /* T /t+ J* >/rit'c> Asicei'iti 


So in Britain all the known facts remain inexplicable unless 

two glacial epochs be recognised. What sort of 

Difficulties explanation, for instance, is given of the inter- 

of the inter- . . . 

mingled arctic mingling of reputed arctic and undoubted tropical 
faunas! PICa an d temperate faunas in the same area and at the 

same time? Prof. Boyd-Dawkins adopts and sup- 
ports with fresh arguments Lyell's desperate "seasonal-migration 

hypothesis," a sort of "Box and Cox" arrangement, 

"Seasonal- ^y wmcn the arctic animals move southwards and 

Migration" occupy the caves in winter, while the tropical move 

northwards and take their place in summer. "In 
the summer the lion, Kafir cat, spotted hyaena and hippopotamus 
would advance northwards ; in the winter the reindeer, musk 
sheep, lemming, tailless hare, glutton and arctic fox would 
swing southwards 1 ." Here it may be asked, were these un- 
wieldy quadrupeds migratory at that time, like birds of passage, 
and could they (think of the hippopotamus) traverse such vast 
distances 2 on foot as these do wearily on the wing? How did 
they manage to hit off arrival and departure so as to avoid com- 
plications between carnivora and herbivora? How were such 
complications avoided on the road between, for instance, fox and 
sheep ? Or are we back again to the nai've days of Noah's Ark 
processions ? And why take the trouble to do all this plodding at 
all, not once but twice a year, merely to fit in with the views of 
naturalists, who for the moment ignore or at least overlook the 
laws of nature and of perspective? For them, peering down the 
grooves of time, warm inter-glacial periods of long duration get 
crumpled up between two glacial epochs brought too closely 
together, and so recede to the vanishing point, leaving everything 
unexplained, and the actual relations involved in chaos. 

The chief trouble appears to be the reindeer, which in all 

Long as- ethnological writings without exception is insepar- 

sociation of a bly associated with an arctic climate, as a foregone 

1 Evidence afforded by the Caves of Great Britain as to the Antiquity of Man , 
Journ. Anthrop. Inst. 1877, p. 156. 

2 To be sure the zones are spoken of as "contiguous" (id.). But the 
respective frontiers alone were contiguous, and large and diversified faunas 
cannot live on frontiers ; they must have elbow-room and spread out over broad 
expanses to avoid congestion; so the difficulty of distances stands. 


conclusion. Yet in these same writings the rein- reindeer and 
deer is habitually preyed upon by the hyaena, so 
that after all the complications are not avoided, 
though other difficulties are created. " In twenty-eight out of 
thirty-one ossiferous caverns the two [spotted hyaena and rein- 
deer] are found side by side, and in the great majority of 
these the gnawed bones and antlers of the reindeer show that 
that animal was the common food of the hyaena" (#. p. 156). 
What then ? Is the hyaena also an arctic animal ? But he was in 
the southern procession. Or was the reindeer a tropical? But 
he was in the northern, and his arctic habitat is taken for granted. 

Is there then no outlet, for the facts are unquestioned? It 
would seem that the foregone conclusion must be reconsidered, 
and indeed put aside. The intimate association of reindeer and 
hyaena shows that the same climate suited both, and that this 
climate was not arctic during the association, but at least tem- 
perate, for few tropical and sub-tropical animals will stand 
great cold, while all the arctic can endure a considerable degree 
of warmth, as in their own short summers, often intensely 
warm. It would further appear that the reindeer was evolved 
like the elk in a temperate or warm zone, and is only arctic by 
pressure and gradual adaptation, so that it has not yet acquired a 
completely variable coat, like the arctic hare and other earlier 
hyperboreans (in winter greyish brown on body, white only on 
neck, belly and hind-quarters). It is identified with Caesar's bos 
cervi figiira (vi. 26), which roamed the Hercynian forest well 
within the historic period, and still lingered in Caithness down to 
the middle of the thirteenth century. Even now it ranges in some 
places as far south as the 5oth parallel, that of the Land's End, 
and those introduced to the London public in 1885 felt perfectly 
at home in the present British climate. 

Almost a more striking case of such adaptation is that of the 
tiger, who infests the Indian jungle and in Java Ana j 
touches the equator. Yet his range extends north- instances, 
wards to the Amur basin and to the island of Great range of 
Sakhalin, " where broken masses of ice have been e lger * 
known to remain heaped up around the eastern headlands till 
the month of July," where the thermometer " often remains 60* F. 
K. 5 


below freezing-point in January," where "icy rains and raging 
snow-storms" prevail "for a great part of the year 1 ." But no 
other tropical animal would appear to show such a range as this, 
without some corresponding special readjustments to the altered 
environment 9 , such, for instance, as that of the white panther 
(Felis isbis) of the Great Pamir 3 , and that of the shaggy-maned 
mammoth surviving till apparently quite recent times in the 
Lena basin and other parts of Siberia. But enough has been said 
to show that there should be little further difficulty about the late 
pliocene and early pleistocene faunas of West and Central Europe. 
It is absolutely unnecessary to assume any such unnatural as- 
sociations as are taken for granted, these apparently incongruous 
interminglings merely bespeaking long mild or warm inter-glacial 
periods, lasting a few hundred thousand years on Croll's com- 
putation, and needed to account for the many secular climatic 
and physiographic changes witnessed by early man in the northern 
hemisphere. It was necessary to dwell on this point, owing to 
the paramount importance attached to the almost universal 
presence of the reindeer in the British and neighbouring palaeo- 
lithic caves, especially by Boyd-Dawkins, who from this fact alone 
concludes against Geikie (Ice Age) that "the perpetual summer 
hypothesis [the warm inter-glacial period] is untenable" (ib. 
p. 161). But in the next chapter it will be seen that in some 
places the reindeer was certainly pre-glacial, and consequently at 
that time belonged to a temperate or warm fauna. 

In respect of the boulder-clays and glacial drift, in which 

palaeolithic objects are constantly found at various 

thfmntrfou^d de P ths on the plateau escarpments and riverside 

in the boulder- terraces, it is to be noticed that their age is 

^ndtenaces* measured, not by the few feet of surface deposits 

accumulated upwards from the level of the finds, 

for which a comparatively short period might suffice, but by the 

extent of the erosions from that level downwards to the present 

level of the river beds. Thus, instead of the slight thickness of 

1 Keane's Reclus^ vol. vi. pp. 453-4. 

2 Even the tiger, however, shows considerable incipient adaptations, the 
northern having longer, softer, and lighter fur than the southern variety. 

3 A specimen of this rare species is now (1895) to be seen in the Paris 
Jardin des Plantes. 


surface deposits on the high-level glacial gravels, we have to take 
into account the 80, 100 or 120 feet of excavations by running 
waters according to the height of the present river-banks. It 
might be supposed that a gauge of the time needed for such 
erosions would be afforded by the extent of the scourings, say, in 
the Thames valley during the historic period (1900 years), or at 
least during the prehistoric (perhaps 2500 or 2600 years) since 
the erection of fortified lines or other works by semi-civilized man 
about the then river level, as at the village of Dorchester below 
Oxford. But the gauge is useless, because, long as the period is, 
it has not had time to operate at all. No appreciable change has 
taken place in the present low level of the Thames even during the 
longer period, that is, since the bronze age. So also in the Somme 
valley, where " there is evidence afforded by relics of the Roman 
and bronze age found in the peat in the bottom of the valley, that 
the river had not materially lowered its bed since those relics were 
deposited, and therefore it must have taken an enormous time to 
work out the whole valley by means' of a river which flowed with 
the same eroding power as at present 1 /' 

It would seem from his division of the glacial epoch (note 
p. 56), that in discussing the antiquity of man Pre ^ Inter . t 
Prof. Prestwich overlooks the two recurrent ice- and Post - 

T .,. ..... ., , glacial man. 

ages. His division into a pre-, mid- and post- 
glacial period leaves no room for an inter-glacial man, which, at 
least in Britain, is the main point at issue. His mid-glacial, or 
simply glacial, is the point of maximum cold, whereas by inter- 
glacial is understood the long warm period between the two ice 
ages. Owing to inattention to this distinction much confusion, 
degenerating into mere logomachy, has been introduced into the 
discussion. Nobody out of Bedlam would suggest that any 
organism, much less the highest, had been differentiated during a 
period of maximum cold with 6000 or 8000 feet of ice on the 
ground. "Yet a "glacial man" is currently spoken of, as if he had 
really made his appearance in such an environment. Well might 
Quinet protest, and refuse to admit that humanity, the loveliest 
flower of creation, " burst into being amid the swamps and fogs 

1 Col. Lane Fox (Genl. Pitt Rivers), Jour. Anthrop. Inst. 1877, P- *78- 


of the glacial epoch V And it may be asked, if palaeolithic 
man was such a "glacial organism,' ' why he should have perished 
in Britain in post-glacial times, or at least before the arrival of 
his neolithic successor, as assumed by Sir John Evans ? Hence 
there is, or ought to be, no question of a glacial or mid-glacial 
man, that is, in the sense of his making his first appearance 
whether by migration or evolution at that untoward moment. If 
he was there, cowering with his huge associates in the caverns 
about the frontal moraines of great glaciers, then he must have 
arrived before the intense cold set in. He did not leave the 
temperate southern lands to cross ice-strewn seas, in quest of such 
precarious shelter as Britain then afforded, at a time also when 
those lands were still connected with the African continent, 
Italy, through Sicily, Malta and Pantellaria; Greece through 
Crete ; Iberia across the straits. 

Hence whoever admits glacial, must perforce admit pre-glacial 
man, for the terms in this connection are practi- 
restate<L blem c^ty synonymous. But pre-glacial may obviously 
be taken in four senses, by which our ideas of man's 
antiquity will be materially affected. It may either coincide with 
the first increase of cold, or with the second, or with the 
warm interval between the two ice ages (inter-glacial), or 
with the still warmer times antecedent to all glacial phenomena, 
stretching back through the pliocene to the miocene when the 
date-palm flourished in Central Europe, and a sub-tropical flora 
stretched far beyond the confines of the Continent in the direction 
of the North Pole. To the first and second the same objections 
must be urged as against glacial man, for all migrations would be 
suspended during periods of steadily increasing cold. Thus the 
question is narrowed down to the inter-glacial between 720,000 and 
240,000 years ago, let us say 500,000, and the indefinite period 
antecedent to the first ice age, say a million of years ago. Of course 
there may be other factors in the problem, such as possible, even 
probable short recrudescences of cold due to local causes in the 
inter-glacial period itself 2 . But such disturbances would not 

1 Quoted by Desor, I? Homme pliocene de la Californie. 

2 Dr Penck, quoted by Rudler and Chisholm, shows that the ice-sheet 
'advanced at least three times over northern Germany as far as Altenburg 


seriously affect the main relations, which would seem on the 
whole to have been as here stated. 

Most authorities now accept an inter-glacial man, who on the 
above computation would have appeared, say, half a 
million years ago, or about a million on the 100 m. 
base (p. 51), or a quarter of a million or less on 
proportionately reduced bases. A pre-glacial man in the broader 
sense, which many cautious writers also concede, and which 
even on the reduced base would allow about half-a-million, might 
be accepted but for the fact that it presupposes fully differentiated 
pliocene Hominidae present in North-west Europe before the first 
ice age. That seems a somewhat violent and unnecessary as- 
sumption, and it was seen (chap, n.) that in pliocene times 
nothing beyond a generalised precursor differing specifically from 
all the present varieties can be looked for, in accordance with 
the general laws of organic evolution. The most rational hypo- 
thesis seems, therefore, that of trier-glacial Hominida, specialised 
not less, possibly much more, than half-a- million years ago. 
This inference derived from the foregoing general considerations 
will be strengthened by a closer study of the primitive Hominidae 
themselves, that is of Pakeolithic Man, in the next chapter. 

Nothing has here been said about a " post-glacial " man, for 
whom so many ardently contend, thinking thereby p os t-giacial 
to limit his term of existence to a few odd thousand ma . n a nonde- 
years. But the term is obviously equivocal, as on 
the assumption of two or more recurrent glacial periods, each 
will necessarily have its post-glacial interlude. Hence a post- 
glacial man may still have a hoary antiquity, or he may have a 
less but still a great age, or he may be comparatively modern, 
according to the particular glacial invasion after which he is 
placed. The term has to be clearly determined, else its discus- 
sion leads to nothing but a war of words, seeing that under 
certain conditions post- and pre-glacial may have precisely the 
same meaning. A post-glacial man following Croll's first ice age 

and Dresden" (Etirope> Stanford Series, p. 27). In the Alps also proof has 
now been obtained of "thu recurrence of at least three successive periods of 
great glaciation, separated by relatively long intervals" (Dr H. R. Mill, Gcog. 
Jottrn., Jan. 1895, p. 68). 


will be pre-glacial to Croll's second ice age ; that is to say, he will 
be inter-glacial, as here maintained. In this conclusion it is 
pleasant to find oneself in accord with such a shrewd and careful 
observer as Enrico Giglioli, who also accepts " inter-glacial" man, 
contemporary of Elephas antiquus and Rhinoceros Merkii^ animals 
at that time characteristic of the Italian fauna 1 . 

1 * * Sappiamo che T Uomo esisteva nei primi tempi del Quaternaries, in quel 
periodo interglaciale per 1' Europa nel quale vivevano nei paesi nostri 1' Elephas 
antiquus ed il Rhinoceros Merkii" (L? Uomo; Sua Autichitti, &c., Florence, 
1893, p. 9). 



Palaeolithic Man spread over the whole world But in many places early and 
later Cultures run in parallel lines, not in time sequence Hence the Time 
relations often obscured, objects of human industry not being everywhere 
tests of age, but only of grades of culture Even these grades not always 
clearly distinguished Paleolithic art not stationary but progressive, and 
in some respects outstripping that of neolithic times Materials available 
for the study of primitive Man: implements, monuments and human 
remains Unreasonable objections to implements (palaeoliths) as evidence 
of antiquity Value of implements determined by their provenance and 
associations in geological formations or in caves Stalagmite beds not 
necessarily a test of age Kitchen-middens of all ages, some very old, 
some recent and of rapid growth, hence to be judged on their merits 
Human remains reserved for special treatment Quaternary Man in 
Britain; Evidence of Hatfield Beds; Kent's Cavern; Brixham Cave; 
Cresswell and Victoria Caves ; Lotherdale and Pont Newydd Caves ; Vale 
of Clwydd Caves ; Thames river-drift ; High-level gravels ; Chalk plateau, 
Kent ; Eoliths from Canterbury gravels, Stoke Newington, &c. Quater- 
nary Man in France : Somme Valley river-drifts, St Acheul Grades of 
Palaeolithic Culture De Mortillet's Four Epochs: Chellian Age, typical 
implements ; Moustierian Age, typical implements ; Solutrian Age, typical 
implements ; Madelenian Age, typical implements The Dordogne School 
of Art Placard Cave : Superimposed Culture eras Evidences of Palaeo- 
lithic Man in France and Italy Quaternary Man in Africa (Egypt, 
Algeria, the Cape) ; in Asia (Syria, Palestine, Asia Minor, Caucasus, Mon- 
golia, India, Japan) ; in Australia and New Zealand ; in America (Tierra 
del Fuego, Patagonia, Argentina and Brazil, Mexico, United States and 
Canada) ; evidence from the Trenton gravels ; Mississippi Basin and other 
localities ; Views of Chamberlin, Holmes, Mason and other conservatives 
on the value of this evidence ; the Calaveras Skull General Diffusion of 
Primitive Man throughout North America The Mound-builders not 
quaternary ; their Culture neolithic, prehistoric and historic. 

IN the last chapter our horizon was mainly confined to the 
northern hemisphere, where the intimate association 
of primitive man with glacial phenomena has been 
most carefully studied. Now it must be extended ^ O c r ^ he whole 
to the southern hemisphere also, thus embracing the 


whole world, which would appear to have been already occupied 
by the Hominidae in palaeolithic times. If this point can be 

established, it will afford of itself a strong argument 

laces eariy any * n SLl PP ort f tne l n g period here claimed for man's 

and later cui- existence on earth. But the question is beset with 

synchronous. snares and pitfalls, due especially to the fact that 

the very terminology itself does not everywhere con- 
note the same order of sequence, much less the same periods of 
absolute time. Thus palaeolithic implements in the New may in 
some cases well correspond with neolithic in the Old World, and 
in all the Continents except Australia, where one order alone 
exists, various phases of progress go on simultaneously rather than 
consecutively. The Aymaras and Peruvians had probably two 
thousand years ago arrived at a somewhat high grade of culture in 
South America, where the Botocudos and the Fuegians have 
scarcely yet reached even the Old Stone Age. Similar contrasts 
are met on the Anahuac tableland, between the long-civilised 
Nahuas and the still barbarous Otomi Highlanders ; in Europe 
between the Slav settlers and their nomad Samoyede neighbours ; 
in Abyssinia between the Semitic Amharas with long historic 
records and the debased Wito fishers of almost aberrant human 
type ; in India between the haughty Caucasic Rajputs and many 
utterly savage Kolarian or Dravidian aborigines. 

Such overiappings of old and new, such persistence of low 

Hence the primitive cultures in the midst of highly advanced 

time relations populations, tend to obscure the time relations, 

often obscured, ^^ are ^ ere un( } er consideration. It is obvious, 

for instance, that implements of the most primitive types, such 
as those of the Tasmanians, more rudely fashioned than those of 
the European palaeolithic hunters, cannot of themselves be any 
test of age. They represent no sequences, but only an incipient 
growth permanently arrested and by adverse conditions prevented 
from attaining its normal development. Where there is no 
change, there is no standard by which to measure time. Hence 
the mistake made especially by some American ethnologists, who 
have assigned a considerable antiquity to certain native cultures, 
solely on the ground of the rude implements with which they 
appeared to be associated. Certain objects, such as flint flakes or 


chippings, if found on or near the surface, or under other circum- 
stances not necessarily involving great age, might have been made 
at any time, and are now still made by many peoples not yet 
brought under higher influences. 

How misleading such evidence may be is shown by the fact 
that, as will presently be seen, some of the European 
palaeolithic are more skilfully worked than the far 

more recent neolithic objects : for the men of the always clearly 

~, , ~ . -,,.,, i distinguished. 

Old Stone Age certainly lived through the greater 
part of the inter-glacial epoch, and they could scarcely have 
remained stationary for such countless generations. They un- 
doubtedly made considerable progress within the limits of their 
primitive culture, and the men of Solutre have left recorded on 
the bones of extinct species various specimens of their artistic 
taste and technique superior to any similar works that can be 
traced to their neolithic successors. 

But this very progress leads to fresh difficulties, as points are 
at last reached where it becomes almost impossible to draw the 
line between the Old and New Stone ages. Although Prof. Boyd- 
Dawkins argues with much force for a great " abyss separating the 
Palaeolithic Age of the Pleistocene period from the Neolithic Age 
of the Prehistoric period," he is fain to admit with Mr J. Allen 
Brown, who takes the continuity view, that the one must be derived 
from the other "in some part of the world"; only we have "not 
yet discovered where that part is; it is probably not in Europe 1 ." 
Yet it would seem from the contents of such caves as those of 
Solutre' and Mentone that even in some parts of Europe there is 
no break but a decided overlapping of the early and later cultures. 
Speaking of the Mentone finds, Mr A. Vaughan Jennings 8 points 
out that while the worked ornaments may be neolithic the 
skeletons, especially those found in 1892, "show osteologicai 
affinities to more ancient types." Even the implements, "though 
not of the type which we in England know as* palaeolithic, are 
certainly not any of the usual neolithic patterns." Hence we 
may here be in the presence of a palaeolithic race in Southern 
Europe, whose culture merges in that of their neolithic successors. 

1 Journ. Anthrop. Inst., Feb. 1894. 

3 The Cave Men of Mentone, Natural Science^ June 1892, p. ay 8. 


Mr Jennings well remarks that "it is likely that the habit of 
speaking of Palaeolithic and Neolithic times, and attempting to 
draw sharp distinctions between them, carries with it more than 
the usual evils of all formal nomenclature " (#.). But this of 
course applies only to the borderlands, where interminglings begin 
to take place between the two cultures, which in other respects 
stand widely apart, and must be studied separately. 

Even the rude bulbed flints themselves already show some 
progress in art. These are generally taken as the beginnings of 
human culture, and Mr Shrubsole 1 and Mr A. M. Bell 1 have 
produced unbulbed palaeoliths from the hill-gravels of Berkshire 
and Mr B. Harrison from the chalk plateau of Kent, which seem 
to be undoubtedly the work of man. They are more advanced 
than those of the South African Bushmen, or the Tasmanians who 
would almost appear to be incapable of fashioning any of these 
British eoliths^ as they are called (see p. 293). Hence any stone 
that can be conveniently grasped must be taken as the true 
starting point, and between this and bulbed flints there is a wide 
interval, with room for much upward development. At the same 
time objects which show no clear sign of artificial treatment are of 
course useless in the study of human progress. 

It is therefore not only the objects themselves, but also the 
associated circumstances that have to be considered. 
Speaking broadly, the materials now available for 

the study of the study of primitive man are threefold, his imple- 

primitive man. /. r ,,.,*,-* 

ments, Jus monuments, and himself. The first, from 
which he rightly takes his name of Palaolithic Man*, are in some 
respects the most important, as being immeasurably the most 
numerous and widespread, but chiefly because they often occur 
under conditions which afford the best proof of their artificers' 
extreme antiquity. The monuments, if such undesigned structures 
as shell-mounds or kitchen-middens may for convenience be so 

1 Jour. Anthrop. Inst., May and August 1894. 

2 Gr. iraXeu6s, old, in the sense of young, early, pristine, and M0os, a stone. 
This convenient term, now universally adopted, is due to Sir John Lubbock, 
who applies it to the first of the four great eras into which he divides Prehistoric 
Archaeology: "This we may call the paleolithic period" (Prehistoric Times, 
p. 2). So neolithic , from P^OJ, new (ib. p. 3). 



named, lie necessarily on the surface, or at most on raised beaches, 
while the fossil remains of man himself have been found almost 
exclusively amid the general contents, or at most under the 
stalagmite floors, of his cave-dwellings. But many of the palaeo- 
liths date from the early pleistocene, while some claim to have 
been discovered in the Suffolk crag (pliocene), and even in the 
miocene of Thenay ; but of this presently. 

Owing to the abundance of materials for the study of early 
man, and one might almost say of man's precursor, 
that have accumulated in every region of the globe, 
especially during the last few decades, a selection 
becomes imperative. In the subjoined summary of evidence of 
the available evidence clearness will be consulted an iqm y ' 
by a geographical, and partly chronological, arrangement of some 
of those objects, whose genuine character seems placed beyond 
reasonable doubt. Even the best authenticated however have been, 
and presumably still are, hotly contested by writers and critics, 
who claim to hold a sort of brief for a narrow orthodoxy which is 
here singularly out of place. It is a remarkable fact that many of 
the pioneers in this line of inquiry have been enlightened Roman 
Catholic or Protestant clergymen, such as Dr Buckland (after- 
wards Dean of Westminster), who in 1821 startled the public by 
the discovery of the remains of no less than seventy-five hygenas 
in the Kirkdale Cavern, Yorkshire, so that it was asked whether 
some antediluvian menagerie had broken loose in those parts. 
He was followed by the Rev. Mr M c Enery, who in 1825 first drew 
attention to the "storehouse of antiquity" preserved beneath the 
stalagmite beds of Kent's Hole ; and the Rev. J. M. Mello, who 
led the way in the exploration of the no less famous Creswell Caves, 
Derbyshire. In France Boucher de Perthes, whose patient re- 
searches amid the high-level drift at St Acheul on the Somme may 
be said to have established quaternary man, was followed by 
the Abbe Bourgeois, who went much further, and whose name will 
always be remembered as one of the ablest champions of tertiary 
man in Europe. When asked how he reconciled such a pro- 
digious antiquity of man with the Mosaic cosmogony, the last- 
mentioned was satisfied to reply : "Je suis naluraliste, je ne fais 
pas de theologie" and overzealous partisans of a forlorn cause 


might remember that nobody need be ptu Papa del Papa stesso. 
Yet some of these writers would have us believe that Sabre-tooth, 
for instance, was still prowling about the Roman castra, or at 
least the early British camping-grounds, and that the hippo- 
potamus was floundering amid the Lincolnshire fens a short time 
before the new era, rather than admit that their associate man 
lived in pleistocene times. " I think," says Mr T. K. Callard, 
"the time has now fairly come to ask calmly the question, 
whether finding the works of man in association with Rhinoceros 
tichorhinns and mammoth, instead of proving man's great antiquity, 
does not rather prove the more recent extinction of these 
mammals 1 "; and again : " The legitimate inference is that he [the 
Woolly Rhinoceros] was contemporaneous with the potters, 
Roman, pre-Roman, or Samian; also that he lived when the 
modern sheep browsed in Creswell dale 1 . 77 Mr Callard is at least 
logical, for he feels that unless the natural history of the Hominidae 
can be made to harmonise altogether with the Mosaic account, a 
few thousand years more or less cannot matter either way ; and 
that is so. 

Here are brought together specimens, so to say, of such 
objects as are indicated at p. 155 in a general way 

Value of im- . . . . 

piements de- as characteristic of palaeolithic times. Their locality, 
theh-^oven- associations and position, together with the names 

of the finders or witnesses, are briefly recorded, so 
tS that the student may be able to judge for himself of 
necessarily a their value as evidence. In doing so it will be 

well to bear in mind three points: (i) that from 
position in or under undisturbed boulder-clays and drift of all 
kinds there is no appeal; such finds must date from pleisto- 
cene times (p. 67); (2) that position in cave-earth under thick 
stalagmite beds does not of itself alone necessarily imply great 
age; stalagmite growth is irregular, as it depends on variable con- 
ditions, amount of rainfall, and quantity of vegetable humus on 
the roof yielding carbonic acid with which the percolating water 
dissolves the limestone particles, thus forming the carbonate of 
lime, which in the cave takes the form of stalactites above and 

1 The Contemporaneity of Man with the Extinct Mammalia, c., pp. 9 
and n. 


stalagmites below ; but normally the growth is slow, and a thick- 
ness of 12 feet, as in Kent's Cavern, may involve . ,_ 

' J Kitchen - 

many thousand years; (3) that position near the middens of 
surface may of itself alone imply great age, as in Hence S must 
old beaches slowly raised to considerable heights be judged on 

, , . , . . their merits. 

above the present sea-level, and in kitchen-middens 

such as some of those in Tierra del Fuego (Elizabeth Island), 
where the shells of which they are composed are extinct, or no 
longer the same as those of the surrounding waters. Kitchen- 
middens themselves cover the whole field from palaeolithic to 
modern times, some being very old, others still in progress, so 
that each has to be taken on its merits. Even size is here no safe 
guide, as well remarked by Mr Petroff: "The time required for 
the formation of a so-called layer of kitchen refuse found under 
the sites of Aleutian or Innuit [Eskimo] dwellings, I am inclined 
to think less than indicated by Mr Ball's calculations. Anybody 
who has watched a healthy Innuit family in the process of making 
a meal on the luscious echinus or sea-urchin, would naturally 
imagine that in the course of a month they might pile up a great 
quantity of spinous debris. Both hands are kept busy conveying 
the sea-fruit to the capacious mouth; with a skilful combined 
action of teeth and tongue, the shell is cracked, the rich contents 
extracted, and the former falls rattling to the ground in a con- 
tinuous shower of fragments until the meal is concluded. A 
family of three or four adults, and perhaps an equal number of 
children, will leave behind them a shell monument of their 
voracity a foot or eighteen inches in height after a single meal.... 
The heaps of refuse created under such circumstances during a 
single season were truly astonishing in size. They will surely 
mislead the ingenious calculator of the antiquities of shell heaps 
a thousand years hence 1 ." 

In consequence of their importance in other connections the 
human remains (skulls, skeletons, whole or frag- 
mentary) are reserved for special treatment. Their 

interest is more than antiquarian : they supply data served for 

,,,., . . , , , . . special treat- 

helpful in determining such fundamental questions ment . 
as the specific unity or diversity of the Hominidse, 

1 Ivan Petroft, American Naturalist, July i88. 


inter-racial resemblances and differences, and the origin and cradle 
of mankind. 


GRAY'S INN LANE, London, pointed implement found at end 
of i yth century, said to have been associated with remains of an 
elephant ; first recorded discovery of a stone object in quaternary 
gravels ; now in British Museum ; Evans' Stone Implements of Gr. 
Britain, p. 522. 

HOXNE, Suffolk, 1797, palseoliths with bones of huge extinct 
Quaternary animals at a great depth in fine brick-earth ; Frere, 

Archceol. xiu. 204. 

HATFIELD BEDS (Brandon, Thetford), East Anglia, post-ter- 
tiary flint-bearing deposits, stratified sands, gravel 
an( * brick-earth underlying boulder-clay of great 
extent and in some districts proved to a thickness 
of 60 feet; T. M c Kenny Hughes, Journ. Anthrop. Inst. 1877, 
pp. 162-65. 

KENT'S CAVERN, one mile from Torquay, Devonshire; 1825- 
94; downward sequence of deposits: i. Limestone 

Kent's Hole. , , , c ,, r - c 

blocks fallen in from the roof, some over too tons, 
partly cemented with carbonate of lime; 2. Black mould 3 to 
12 in. thick with remains of living species only, also Roman and 
pre-Roman objects (potsherds, &c.); 3. Cave-earth and black band 
4 ft. thick underlying granular stalagmite 5 ft. thick, with char- 
coal, burnt bones, 366 flints delicately made of flakes but never 
polished, also needle and other bone objects, but no pottery; 
living and extinct faunas, hyaena, mammoth, cave-bear, horse, 
glutton, cave-lion, reindeer, rhinoceros tichorhinus, urus, machai- 
rodus latidens, voles, Irish deer, hare; 4. Breccia derived from 
neighbouring hills underlying crystalline stalagmite nearly 12 ft. 
thick, with rude massive implements made of flint nodules, but 
also a flint flake and a chip embedded in the breccia; fauna 
chiefly ursine, with lion and fox; inscriptions or graffiti in cave 
with dates 1604, 1615 and 1688, the oldest with thin stalagmite 
accretion showing rate of growth about -fa inch in 250 years. 
Buckland, M c Enery, Pengelly (numerous writings), Ralph Richard- 
son (Transactions Edinburgh GeoL Soc. 1886-87). 




a. Flint implement ; b. bone awl j c. harpoon head ; d. needle. 





BRIXHAM CAVE, near Kent's Cavern, with similar contents ; 
Pengelly (Iteport of Committee of the Royal and GeoL 

TREMEIRCHION CAVE, Vale of Civvy d, N. Wales, 1885; flint 
lance-head and scraper with cave-lion, mammoth, woolly rhinoceros, 
hyoena, &c. ; Hicks and Davies, Quart. Jour n. GeoL Soc. XLII., p. 3. 
CRESWELL CAVES (Robinhood, Church Hole and Pin-Hole), 
Creswell Crags, Derbyshire, 1875-6; upper bed 
under stalagmite up to i ft. thick; quartzite, flint, 
ironstone tools, scrapers, spear-heads, bone awls, 


RIVER DRIFT IMPLEMENT (Paleolithic}. From Saiitou Downham, 


borers, rib-bone with incised head of hog-mained horse (objection 
to this that it implied clipping and palaeolithic shears invalid, 
because hog-mane is a natural growth) ; machairodus, lion, 
leopard, hysena, reindeer, woolly rhinoceros, hippopotamus, Irish 
deer; lowest bed 3 ft. thick, red sand and clay; rude quartzite 
flakes; lion, reindeer, hyaena, mammoth, rhinoceros, Irish deer, 
horse, bear, wolf; tools of this bed identical with those of 
Brandon, Bedford, Hoxne, St Acheul and thence south to 
Toulouse, always associated with reindeer, mammoth, woolly 
rhinoceros, &c. 


VICTORIA CAVE, Settle, West Riding, Yorkshire, on the 
Kibble; cave so called because opened 1837, year 
of Queen Victoria's accession; "hyaena bed" under C ave? r * 
glacial deposits originally 25 now 15 ft. thick; 
scratched bones, also human or ursine fibula (Burk's "bone of con- 
tention"); elephas antiquus, mammoth, hippopotamus, bos primi- 
genius, rhinoceros leptorhinus, bear; Rev. J. M. Mello, Boyd- 
Dawkins, Tiddeman, Crosskey; the fibula doubtful, but the 
bones found 1875 an d 1876 show clean cuts or markings of 
human agency ; one is small humerus of goat generally supposed 
to be a late arrival coming in with the neolithic herdsmen, but 
shown by M. E. Dupont to have associated with rhinoceros, 
hippopotamus, cave-bear and other extinct fauna (L Horn me pen- 
dant les ages de la Pierre, p. 197, and letter to Tiddeman in Jour. 
Anthrop. Inst. 1877, p. 168) ; the hyaena bed is certainly pre- 
glacial in N. Britain, which may correspond to a post-glacial period 
in the south, there being evidence of two strongly-marked glacial 
periods, the earlier reaching far south, the latter arrested in the 
north of the Midland Counties (Tiddeman). 

LOTHERDALE CAVE, near Skipton, same fauna as in Victoria 
Cave in old river-gravel bed under glacial deposits, 

- i / i n i j Lotherdale 

but not in the river-gravels of the well-glaciated sur- and Pont 
rounding district, showing that here the reindeer caves'** 
did not come in with, but preceded the ice-age. 

PONT NEWYDD CAVE, Denbighshire, 3 miles from St Asaph, 
flints and associated extinct fauna in p re-glacial deposits; see p. 
63 ; here are remains of hippopotamus, rhinoceros hemitcechus 
and elephas antiquus. 

Clwyd, explored 1884-86 by Dr. H. Hicks and 
Mr E. B. Luxmore ; occupied by pleistocene animals C1 ^ Caves, 
and man before deposit of the surrounding glacial 
beds ; the caves now 400 ft. above sea-level, yet the contents had 
been disturbed (remanie^ resorted) by marine action; above the 
fossil-bearing beds were deposits with foreign pebbles like those 
of the glacial beds; Stet had been blocked by thick glacial beds 
necessarily deposited after its occupation by the pleistocene 
fauna; a small well- worked flint flake in the bone-earth (18 inches 
K. 6 


below the lowest layer of sand) which extends outwards from the 
entrance under the glacial beds, which were proved to over 2 oft. 
The lance-heads, scrapers and other implements appear to be all of 
same age as the flint flake, and it is evident that the contents (of 
Stet) had been washed out by marine action during submergence 
in mid-glacial times and then covered by marine sand and an 
upper boulder clay (Hicks, paper Brit. Ass. Meeting, 1886). 


Numerous palseoliths of normal river-drift type; 

river?drift early pleistocene ; spotted hysena, lion, &c.; 26 

species, of which six only extinct ; Rev. Osmund 

Fisher; Cheadle and Woodward (Boyd-Dawkins, Nature, Aug. 31, 

1882, p. 436). 

FINCHAMPSTEAD HILL-GRAVELS, Berkshire, pre-glacial and 
apparently pre-pleistocene, deposited by a river 
tnat nas ceased to exist; extremely rude palaeo- 
liths ("eoliths 1 "), grooved scrapers, large imple- 
ments with rounded butt, flints worked at poStt only; figured by 
O. A. Shrubsole \i\four. Anthrop. Inst. Augus|Pi894, p. 44. 

.CHALK PLATEAU, Kent, rolled and other rude palaeoliths, 
chalk like the Berkshire eoliths, described by A. M. 

Plateau. Bell,/0*r. Anthrop. Inst. May, 1894, p. 266 etseg. 

TON, &c. &c. Numerous eoliths and palaeoliths of 
a ^ types an d forms collected and described by 
Mr Worthington G. Smith in Man, the Primeval 
Savage, 1894. Many of these are so rude that they may well be 
assigned to a tertiary or eolithic precursor whenever his existence 
is established in Britain. But "on this question the world needs 
enlightenment " (Thos. Wilson, Prehistoric Anthropology ', p. 604). 

European Mainland. 

ST ACHEUL, ABBEVILLE, AMIENS. Somme Valley, explored 

Ouatcrna * F man Y y ears ( I ^4i i86o),byBoucher de Perthes, 

man in who found numerous palseoliths of ordinary types 

associated with extinct fauna in undisturbed high- 

1 Or. 770;?, dawn, and \ldos, a stone. 


level river-gravels; sites visited (1859-60) by Prestwick, Evans, 
Lyell, Flower, and other eminent English scientists, who verified 
Boucher's statements as "established beyond all controversy" 
(Evans, The Progress of Archaology, 1891, p. 5, and else- 
where). The acceptance of palaeolithic man on the mainland 
by competent judges beyond suspicion dates from this event. 
Boucher's first actual find was a rudely fashioned flint in a 
sandbank at Menchecourt, 1841. Further research has enabled 
archaeologists to divide me palaeolithic age more or less satis- 
factorily into various epochs or sequences according to the 
faunas associated with the implements or the localities where 
found. Thus M. Lartet makes three such divisions, those of 
the cave-bear, mammoth and reindeer, reduced by _ . - 

' tirades of 

Dupont to two, mammoth and reindeer. These Palaeolithic 
cannot be accepted because of the intermingling Culture - 
of the faunas (p. 64) J . Evans, followed by Cartailhac, Reinach 
and others, proposes two, the alluvium, and the caverns, as if 
primitive man first occupied the land, and was then driven by 
the increasing cold to take shelter in the caves. But from their 
contents it is evident that the caves were inhabited at all times ; 
and there is no reason why that should not be so. 

At present the most generally accepted and perhaps the most 
convenient division is that of M. de Mortillet 2 into 
four epochs, or culture sequences, named from the four Epochs?* S 
places in France where the most numerous and 

1 The zoological divisions, as they may be called, have led to endless con- 
fusion and misunderstandings, as when de Quatrefages argues that the men of 
Furfooz (Lesse Valley, Dinant), must have been palaeolithic merely because of 
their association with the reindeer, the lemming and a few other animals 
assumed to belong necessarily to an Arctic fauna {Races Hnmaiiies, Questions 
Generates, p. 74). Yet these men made pottery, deposited offerings with the 
dead, and were sub-brachycephalic, a combination of characters indicating a 
distinctly neolithic race. No doubt pottery was also found by Dupont at the 
Trou <ie Chaleux> same district, under debris over 3 ft. thick that had fallen in 
from the roof; but that is no test at all of age. Such accidents may happen in 
a moment at any time, and objects found under such debris in Kent's Cavern, 
Placard and elsewhere, are often not only neolithic but even historical (British, 
Roman &c.). The rudest pottery has not yet been traced to distinctly pleisto- 
cene and inter-glacial times in Europe. 

2 In his classical work, Le Prthistorique. 



most typical implements of the several epochs have been found. 
These are the Chellian, from Chelles a few miles east of Paris; 
the Moustierian from the cave of Moustier on the river Ve'zere, 
Dordogne; the Sohitrian, from the cave at Solutre near Macon; 
and the Madelenian, from the rocky shelter of La Madeleine, Dor- 
dogne. This nomenclature is of course purely conventional, the 
local names being taken merely as indicating so many types, to 
which implements have to be referred wherever found. The chief 
objection is perhaps the fact that the human remains from Solutre 
appear to be not palaeolithic but neolithic. At least the skulls 
are not dolicho- but brachy-cephalic, and "hitherto no certain 
example of brachycephaly has been found amongst quaternary 
human remains" (Salmon, ib. p. 6). But this question must 
not be prejudged, and meanwhile de Mortillet's fourfold division 
appears to hold the field. In any case it has been too widely 
accepted 1 to be overlooked in any comprehensive ethnological 
treatise. It takes no account of an eolithic period, which has 
yet to be established; nor is it probable that the grouping will 
be found elastic enough to meet all cases with the progress of 
discovery in every part of the world. It must therefore be re- 
garded, not as possessing finality, but only as a convenient 
scaffolding, to be removed when it has served its purpose, that 
is, when the last word has been said on the obscure problem 
of palaeolithic man, his age, evolution and general culture. 

CHELLES, right bank Marne, district Meaux, above Paris; 
numerous chipped flints, mostly oval or almond- 

Chellianage. rr ' J 

shaped, some more round and even like dirks 
(scrapers?), cutting edge generally at the point, but also ex- 
tended nearly round, leaving part for a grip. "I much doubt 
whether any of them were attached to a handle" (Wilson, ib. 
p. 608). Since their manufacture many have been deeply patined 
and rusted sometimes even right through, in red, yellow, or chalky 
white colours by physical or chemical agency, implying great age; 
uses obviously multifarious at a time when this was almost the only 
implement invented by man; this "Chellian type" is found almost 

1 To mention one instance, de Mortillet's division has been taken as 
the basis of Mr Thomas Wilson's excellent Study of Prehistoric Anthropology* 
Washington, 1890. 



From Redhill, Thetford. 


everywhere in both hemispheres as far north as the "Arctic Circle" 
which at that time included Scotland, Scandinavia, North Bel- 
gium, N. Holland, N. Germany and N. Russia; it answers 
generally to the river-drift type of Britain. In the St Germain 
Museum, Paris, are six cases of these flints from the Chelles 
sands and gravels, which rest to a thickness of 22 to 26 ft. on 
the original chalk, some of the objects being coeval with the first 
deposits, consequently of vast age. A great part of France is 
covered with the plateau formation through which the rivers have 
eaten their way down to their present levels. In this formation 
multitudes of flints of the Chellian type are found ; consequently 
at that time, before the running waters had begun their erosive 
work, man lived in relatively numerous communities more in the 
open air than in caves; hence the climate was mild and either 
pre-glacial absolutely (late pliocene), or inter-glacial between 
CrolPs two great ice-ages (pleistocene). Either assumption an- 
swers the conditions, though the absence of Chellian flints from 
the then Arctic regions (see above) would imply that the period 
was rather inter- than pre-glacial; otherwise there seems no 
reason why primitive man should not have ranged northwards, to 
Scandinavia for instance, at a time when the climate of high 
latitudes was favourable. But no true Chellian or other palaeo- 
lithic implements occur in Scandinavia, despite the statements 
of Zinck and others to the contrary (Wilson, ib. pp. 74-5). 
After the Chellian follow what may be called the Cavern periods, 
that is times when man resorted more to the caves than to the 
open, as if the first ice-age were now setting in. It will be con- 
venient to keep these periods, as named by de Mortillet, together, 
as under : 

MOUSTIER CAVE, on the right bank of the Vzere affluent of 
the Dordogne, above Les Eyzies and Tayac ; typical 
* 11 implements, flint point or spear-head left smooth 

mints Imple " anc * ^ at on one s ^ e > as struc k fr m tne core, 
pointed and edged from the other side; scraper 
treated in same way, but with edge rather upon the side than 
at the end, as in all succeeding epochs ; similar objects occur in 
the river-gravels, but are found in the caves at such depths and in 
such associations as to suggest long occupation during glacial 


times with a fauna more like the present, all the now extinct 
forms having already disappeared. In fact some of these flint 
implements, which from their form are treated by the French 
geologists as palaeolithic, "would be included in the second 
division, or neolithic, in England 1 ." 

SOLUTR& CAVE, Macon district, Saone-et-Loire; flint imple- 
ments of laurel-leaf and other patterns, showing an 

, i r i 11 Solutrian or 

immense advance on those of the previous age, both second Cave 
in variety of form and especially in finish, in this ments lmple " 
respect scarcely ever since rivalled, certainly never 
surpassed ; large thin spear-heads ; scrapers with edge no longer 
on the side but on the end; flint knives and saws, but all still 
chipped, never ground or polished; characteristic are the long 
spear-points with tang and shoulder on one side only ; also bone 
or horn awls or borers. These beautiful objects occur in nests or 
caches as in the United States. The flakes chipped off are some- 
times "so long and thin as to resemble shavings rather than 
chips" (Wilson, p. 615). The fine laurel-leaf patterns defy imita- 
tion, hence have never been forged like so many other antiques. 
Besides Solutre' they occur in several other caves, such as Rigny- 
sur-Arroux (Saone-et-Loire), Grotte de Garges (Vaucluse), and 
Grotte de TEglise (Dordogne). 

LA MADELEINE ROCK SHELTER, on the Vezere, about mid- 
way between Moustier and Les Eyzies ; a very long Madeienian 
epoch represented by numerous stations, whose or Third Cave 
varied contents show continued progress in the arts ge * 
and general culture; scrapers, gravers, saws and knives of flint 

1 J. Allen Brown, Jour. Anthrop. Inst. 1893, p. 92. This palaeontologist, 
it may be mentioned, proposes (ib. p. 94) four divisions instead of the two 
commonly accepted: (i) EolUhic, roughly hewn pebbles, nodules &c., of the 
chalk plateaux older than the present hydrographic system; (2) Paleolithic \ 
flints of the higher river drift of the present valleys, and oldest limestone 
cave breccias; (3) Mesolithic* flints of better form intermediate between the 
palaeolithic and (4) Neolithic^ polished or delicately worked implements like 
those from the Danish tumuli, dolmens, &c. These divisions will probably be 
accepted when sufficient data have been collected and correlated to clearly dis- 
tinguish between the several epochs. But even then there will always be inter- 
minglings and overlappings. 




flakes; borers, needles, "harpoons," hooks and diverse orna- 
ments of bone, horn and ivory; but specially re- 
pigments! im " markable are the spirited carvings in the round, or 
etchings on stone, bone or horn, of seals, fishes, rein- 
deer, mammoths and other animals, including man himself, besides 
decorative work in straight, curved, or dotted lines, zigzags, 
festoons or herring-bone pattern. This palaeolithic " school of 
art" stands apart, being almost exclusively confined to the 
Dordogne district, although, as seen, analogous specimens occur 
in Britain (Creswell), also in Belgium (Cave of Goyet), and a few 
other places. Besides La Madeleine, the chief stations of this 
epoch are Les Eyzies, Laugerie Basse and Gorge d'Enfer, in 
Dordogne; Grotte du Placard, in Charente; and others in South- 
West France. Noteworthy are a mammoth engraved on a frag- 


a. Of pike, cut on canine tooth of bear, Duruthy Cave. 

b. Of aurochs, trees, snake etc. on reindeer horn, La Madeleine. 


ment of its own ivory tusk, a dagger of reindeer horn with handle 
in form of a reindeer, a cave-bear incised on a flat piece of schist, 
a seal on a bear's canine, a fish admirably carved on reindeer 
horn, and a scene also on reindeer horn, showing horses, aurochs, 
trees, and a snake biting a man's leg. The man is naked, and 
this with the horses and snake 1 suggests a warm climate, despite 
the " Arctic " reindeer. Horses as well as reindeer abounded in 
this and the previous epoch, and the Solutre' cave district alone 
yielded the fossil remains of about 10,000 of these animals. 
Sufficient attention has not been paid to such points by those 
ethnologists who regard all the cave men as "glacial." For the 
*' arctic reindeer epoch " of many French writers might be sub- 
stituted a " temperate horse epoch " more in harmony with the 
prevailing relations. In many instances art is displayed for its 
own sake, as in the embellishment of the so-called "batons de 
commandement," apparently a kind of mace or emblem of 
authority, from the Madeleine and Goyet caves. But this culture 
suffered a sudden eclipse either before, or coincidently with, the 
irruption of the rude neolithic peoples into Western Europe, just 
as on the Anahuac plateau the Toltec culture disappeared before 
the invasion of the Chichimec barbarians, and was not again 
revived till two or three centuries before the arrival of the Con- 

PLACARD CAVE, on the Tardoire affluent of the Charente river, 
is the "Kent's Cavern " of France, its several layers 
revealing like it the successive phases of troglodytic cave, 
culture from the Moustierian upwards. In the 
accompanying cut is seen the grotto with a sec- 
tional view drawn to scale of the several beds, which in de- 
scending order are as under : 

AAA. Debris fallen from roof at various periods and separating 
the different beds ; represents the stalagmite floors of limestone 
caves. B. Layer of same with thin streak of clay interposed ; 
no remains in A or B. C. Top implement-bearing beds 15 in. 
thick ; polished flint hatchets, barbed spear-heads, bones of living 
species, all neolithic. D. E. F. G. All with Madelenian objects 

1 It has been called an "eel"; but the action shows that it is clearly a 



and corresponding species; altogether, with intervening strata, 
about loi feet thick. H. Late Solutrian beds ; spear-heads with 


shoulder on one side. I. Earlier Solutrian bed with laurel-leaf flints. 
K. Moustierian bed, with typical pointed flint. 

Note that the second layer A (debris from 
roof) is 28 in. thick, so far implying non-con- 
tinuity and a considerable gap between the latest 
paleolithic (D) and the neolithic (C) era. Similar 
evidence is afforded by other caves, such as 
Laugerie Haute (gap 50 in.), and Grotte de la 
Vache (stalagmite gap 21 in.). But these gaps, 
by which many eminent archaeologists have been 
influenced, do not necessarily imply correspond- 
ing breaks between palaeo- and neo-lithic times. 
They merely point at gradual improvement in 
the climate after the retreat of the last ice-sheet, 
thus enabling the men of the New Stone Age to 
live more in the open, and obliging them to 
resort less frequently to the caves, which were 
at last abandoned altogether. In other words, 
early neolithic man was of less troglodytic habits 
than his immediate Madelenian precursor of 




post-glacial times. So far it may be inferred that in Britain 
and West Europe primitive man appeared after the retreat of 
CrolFs first ice-sheet, and lived throughout the whole of the 
inter-glacial period, surviving in some places till the retreat of 
Croll's second ice-sheet, when he was gradually replaced and 
no doubt partly absorbed by early neolithic man arriving after the 
final disappearance of glaciation everywhere below the Alpine 

THENAY, near Pontleroy, Loir-et-Cher, flints extracted by 
TAbbe* Bourgeois from the miocene beds (knife, 
scraper, point) ; claimed by him to be of human Evidences of 
workmanship, but claim generally disallowed, and in France, 
doubted by dc Quatrefages (ib. p. 93). These finds ? gal and 
first raised the question of tertiary man in Europe, 
further proof of which was adduced in 1863 by M. Desnoyers, 
who produced from the gravels of 

SAINT-PREST, near Chartres, various incised bones, undoubtedly 
worked by man and associated with elephas meridionalis and 
rhinoceros leptorhinus ; site examined by Lyell, but the beds 
appear to be rather old quaternary than true tertiary. Better 
evidence was brought forward by M. Rames from the upper 
miocene of 

Puv-CoURNY, near Aurillac, Cantal; by Senhor Ribeiro from 
the same formation at 

OTTA, Tagus Valley, near Lisbon ; and by Signer Capellini 
from the pliocene of 

MONTE-APERTO, near Siena. Apart from human remains, 
the proofs of tertiary man in Europe appear at present to be 
limited to these finds, the most convincing of which are those ot 
Rames, of some of which it is admitted that "had they been 
found in quaternary beds no one would have hesitated to regard 
them as intentionally carved " (de Quatrefages, ib. p. 93). This 
carries the question back, not merely to a pliocene, but to a 
miocene (mid-tertiary) precursor of the Hominidae. Such a 
precursor would be necessarily intermediate between the gene- 
ralised pliocene precursor, who must be accepted, and some 
higher anthropoid forms than any of those now existing. Such a 
being must no doubt also be postulated ; but he could scarcely 


be regarded as distinctly human, though possibly endowed with 
sufficient intelligence to work the rude flints produced by M. 


NILE VALLEY, opposite Thebes, chert implements from the 
undisturbed river-drift, and from the breccia in 
which are hewn the royal tombs of the wady Biban Manlrf /UKca 
el-Moliik; Pitt Rivers, Journ. Anthrop. Inst., June 

ABYDOS DISTRICT, 30 miles north of Thebes, a limestone 
plateau 1400 feet above the present Nile level, explored (1894- 
95) by Prof. Flinders Petrie and described by him as " the home 
of palaeolithic man." Here were found in great numbers large 
massive flints, beautifully worked and perfectly unworn, "of 
exactly the same forms as those of France and England. The 
enormous age of these is shown by their black-brown staining, 
while others 5000 years old by their side show scarcely a tinge of 
weathering. Besides these, other flints of a later palaeolithic 
type are found embedded in the ancient gravels of the former 
High Nile, so that the Nile still rolled down as a vast torrent 
fifty times its present volume at the latter age of palaeolithic 
man" (Paper read before the Royal Society, Edinburgh, April, 

CAIRO, fine river-drift hatchet found 1879 on the road to 
the petrified forest 3 miles from the city, made to be grasped, 
not fixed to a handle (H. Stopes, Rep. Brit. Ass., 1880, 
p. 624). 

GAFSA and surrounding district, Tunisia; successive epochs 
of palaeolithic culture (Chellian, Moustierian and Solutrian) ; work- 
shops and immense numbers of typical implements, some found 
in the undisturbed gravels 16 ft. below the surface, as on the 
right bank of the Wed-Baiash ij mile north of Sidi-Mansiir ; no 
pottery or polished stones, but fragments of friable bones carved 
with fine parallel lines, and one with the rude outlines of an 
animal's head. The long sojourn of palaeolithic man in Tunisia 
west from Gulf of Cabes (Syrtis Minor) is placed beyond all 


doubt by the explorations of Dr R. Collignon 1 and Dr Couil- 
lault 2 . 

TLEMCEN, near Oran; KOLEA, west of Algiers, and other 
parts of Algeria numerous flints of distinct palaeolithic type, but 
mostly surface finds; Lubbock, Bleicker, Hayness. 

CAPE COLONY, NATAL, stone implements from every part of 
this region, some (Natal) undoubtedly palaeolithic of river-drift 
types (knives, scrapers, spear-heads, &c.); W. D. Cooch (illus- 
trated memoir), J. Sanderson. 


SYRIA, palaeolithic hatchet, found 1842 by the Abbe* Richard 
between Mt Tabor and Sea of Tiberias. 

PALESTINE, another of same type found 1880 between Jeru- 
salem and Bethlehem by H. Stopes {Antiquity of 
Man, p. 7). "This axe has been chipped and ManVnAsY^ 
worn in use, and the chips have during the vast 
lapse of time it has been exposed to the weather assumed that 
peculiar appearance that lengthened exposure alone gives" (#.). 

LEBANON, quaternary station with palaeolithic tools associated 
with partly extinct fauna (Louis Lartet). 

ASIA MINOR, hatchet of river-drift type from Abydos (Lub- 

CAUCASUS, cave 30 miles from Kutais, human remains with cave- 
bear and other large fauna (Prince Mossa Shvili, M. Navrotsky). 

MONGOLIA, arrow-heads from quaternary beds near Tul-she- 
san-hao (Abbe* Armand David). 

INDIA, numerous palseoliths from pleistocene beds in every 
part of the peninsula, generally of same types as the European 
river-drift; some near Madras under thick beds of laterite (Med- 
licott and Blandford); quartzite hatchet from the fossiliferous 
undisturbed beds of the Narbadda (Racket); agate knife from 

1 Les Ages de la pierre en Tunisia in Matlriaux pour fhistoire primitive et 
naturelledeVhomme, $rd series, Vol. iv., May, 1887. 

2 Stations prthistoriques de Gafsa (Tunisie), in L? Anthropologit, V. 5, 
1894, p. 530 et seq. "On peut done en conclure que les populations primitives 
qui taillaient ces silex ont ete tres re*pandues dans toute cette region du Sud 
tunisien, a 1'epoque oil s'operait le lent comblement des valle*es" (ib. p. 533). 


corresponding pleistocene beds of the Godavery (Winne); both 
associated with hippopotamus, elephas insignis and other large 
extinct (pliocene?) fauna. 

JAPAN abounds in caves and shell-mounds of great age, 
studied by John Milne (Stone Age in Japan, Journ. Anthrop. 
Inst., May, 1881); but not yet brought into clear relation with 
pleistocene times in Europe ; of the caves there are vast numbers, 
many opening southwards and supposed to be artificial. "It is 
more than probable that they offer as wide a field for the research 
of the cave-hunter as caves do in any other country, and from 
them a rich harvest of facts relating to prehistoric times has yet 
to be reaped." The shell heaps of Nemuro, Hakodate, Omori 
near Tokyo and others, stand 20 or 30 ft. above the present sea- 
level, and those of Omori lie about half a mile from the present 
shore line (ib. p. 414). 

AUSTRALIA, numerous mirrnyongs (ash-heaps, shell-mounds, &c.) 
mainly confined to the eastern and southern regions ; 

Quaternary ] ,.,,,. 

Man in some very large and evidently of great age; one 

Australia. Q tway 3OQ x ^ ft 

must have taken ages for the fish-eating natives of the coast to 
build up such heaps" (Brough Smyth, Aborigines of Victoria, n. 
p. 234, and Vol. i. p. 239): "the layers of which they are com- 
posed point clearly to the slow and gradual heaping-up of small 
quantities of material from time to time." Near the south-west 
coast between Port George the Fourth and Hanover Bay there is 
"a complete hill of broken shells, which it must have taken some 
centuries to form, for it covered nearly, if not quite, half an acre 
of ground, and in some places was 10 ft. high" (Grey, North- 
Western and Western Australia, I. p. no). E. M. Curr, who has 
examined a great many of these "ovens," states that "neither 
stone arrow-heads nor fragments of pottery are found in them" 
(The Australian Race, in. p. 677). But the "stone-circles" men- 
tioned in Chambers* Monuments of Unrecorded Ages as "numerous 
in Victoria," have no existence; "there are no such circles, and 
never were" (Smyth, n. p. 235). On the other hand vast 
numbers of stone implements (hatchets, knives, adzes, scrapers, 
pounders, points, &c.), made of diorite, basalt, quartzite, granite, 
porphyry, obsidian, lava, sandstones, &c.), occur almost every- 


where, but always on or near the surface. " It is scarcely possible 
to disturb any large area of the natural surface in Victoria without 
lighting on some of these weapons.... Broken tomahawks, broken 
adzes, chips and flakes of basalt, and near the coast old mirrn- 
yong heaps, which for ages have been covered with drift-sand 
[blown sand] are from time to time discovered. All these show 
that the Aboriginals, living in exactly the same state as they were 
found when Australia was first discovered, have been for periods in- 
calculable the possessors of the soil... But though some hundreds 
of square miles of alluvia have been turned over in mining for 
gold, not a trace of any work of human hands has been dis- 
covered. Some of the drifts are not more than three or four feet 
in thickness (from the surface to the bed rock), and the fact that 
no Aboriginal implement, no bone belonging to man, has been 
met with, is startling and perplexing" (Smyth, i. p. 364). And 
although some implements are chipped, others ground and polished, 
the distinction is rather one of locality and material than of age. 
" There is no method by which we can distinguish a difference of 
period if we examine stone implements " (fb. p. 360) l . Neverthe- 
less a strong proof of vast antiquity answering perhaps to that of 
palaeolithic man in the northern hemisphere, is afforded not only 
by all this cumulative evidence, but also "by the fact that in 
sinking wells and other excavations in the Hunter Valley, flat 
rocks with axe-marks on their surfaces have been discovered at 
the depth of 30 feet or more below the present surface-level, and 
covered with drift or alluvium, which in all probability must 
have taken thousands of years to accumulate" (Bennett, History 
of Australian Discovery and Colonization^ p. 263, quoted by 

In NEW ZEALAND three stone ages are distinguished by 
F. R. Chapman, the last being the contemporary, associated 

1 So also R. Etheridge: Has man a Geological History in Australia? Froc. 
Linn. Soc., N. S. Wales, 1890, p. 259; and E. H. Giglioli: Le Eta della 
Pietra net? Australasia, &c., 1894, p. 4 : "II tipo e la fattura delle arme e 
degli strumenti di pietra degli Australian!, e dico ci6 per osservazioni mie 
proprie su un esteso materiale nella mia collezione, rappresentano oggi tutti gli 
stadii possibili dal piti rozzo tipo paleolitico al saggio piti perfezionato...deir 
epoca neolitica." 


with the working of nephrite or greenstone by the Maoris (Trans. 
N. Zealand Inst.) 1891, p. 479). The first may be referred to the 
Papuan predecessors of the Maori, the " Moa-hunters," although 
the ancestors of the Maori themselves appear to have hunted 
some species of dinornis. In any case its correlation to the 
extinct pleistocene fauna of the northern hemisphere remains to 
be determined. 


TIERRA DEL FuEGO \ kitchen-middens ancient and modern, 
both of prodigious extent, and formerly much larger, 

Quaternary r & J 

Man in having suffered greatly from marine erosions. The 

menca. former, after every allowance is made for rapid 

accumulation, are shown from their contents and magnitude 
to be of vast age, and considering their position at the southern 
extremity of the Continent, seem alone sufficient to solve in the 
affirmative the question of quaternary man in the New World. 
What remains of the shell-heap on Elizabeth Island is nearly a 
mile long, stands 24 ft. above the present sea-level, has a mean 
thickness of nearly 4 ft,, and is covered with a layer of fine sand 
from 24 to 28 in. thick, above which is a layer of vegetal )le 
humus with a luxuriant herbaceous growth. Lovisato, who has 
carefully studied these Fuegian shell-mounds, shows that that of 
Elizabeth Island was submerged, during submergence received 
its layer of marine sands, and was then upheaved to its present 
level. He also shows that the shells (patella, mytilus) forming a 
great part of the contents are different from and much larger than 
the corresponding species now inhabiting the surrounding waters 1 . 
Similar phenomena are presented by the mound at Ushwaya in 
Beagle Channel, and by the other ancient middens strewn over 
the Archipelago. 

PATAGONIA. Here as in so many other parts of the New 

1 "Le valve delle grandi patelle e le altre del grossi mitili del deposito non 
si trovano oggigiorno su quelle spiaggie, ne sulle circostanti, ove patelle e 
mitili, che pur vivono ancor in quel mare sebbene non abbondanti, sono pic- 
colissimi " (op. cit. p. 1 1). This argument will appeal forcibly to those 
palaeontologists, who are well aware how very slow is especially the growth 
and evolution of these organisms. 


World the great difficulty arises, not from lack of material, of 
which there is a superabundance, but from the intermingling or 
close juxtaposition of types, the persistence of old in the midst of 
new forms, so that it often becomes impossible to discriminate 
between remote and later epochs. In the Western Hemisphere 
there are few Creswell or Placard Caves, where the relics of the 
past follow in orderly succession, as if arranged in cabinets for the 
convenience of the antiquarian student. Thus the Rio Negro 
Valley, Patagonia, may rather be compared to an ill-assorted 
ethnological museum, where the naturalist, Mr W. H. Hudson, 
wanders about the abandoned sites of old and recent habitations 
profusely strewn mostly on or near the surface with evidences of 
the presence of primitive and later generations. Nevertheless, 
thanks to denudation and weathering here and there, "the sites 
of numberless villages 1 of the former inhabitants of the valley 
have been brought to light. I have visited a dozen such village 
sites in the course of one hour's walk, so numerous were they. 
Where the village had been a populous one, or inhabited for 
a long period, the ground was a perfect bed of chipped stones, 
and among these fragments were found arrow-heads, flint knives 
and scrapers, mortnrs and pestles, large round stones with a 
groove in the middle, pieces of large polished stones used as 
anvils, perforated shells, fragments of pottery, and bones of 

animals The arrow-heads were of two widely different kinds 

the large and rudely fashioned, resembling the palaeolithic 
arrow-heads of Europe, and the highly-finished or neolithic, of 
various forms and sizes. Here there were the remains of 
the two great periods of the Stone Age, the last of which 
continued down till the discovery and colonization of the 
country by Europeans. The weapons and other objects of the 
latter period were the most abundant, and occurred in the valley -, 
the ruder were found on the hill-sides, in places where the river 

1 Such sites, the paraderos of the Hispano- American writers, are scattered 
in great numbers all over Argentina. Although the contents are mostly those 
of neolithic times, some, such as that of the Marco-Diaz Valley (612 x 408 feet), 
must have been occupied either continuously or at intervals for untold gene- 
rations. Paradero is the Spanish "sojourn," "residence," from parar, to 
stop, or sojourn. 

K. 7 


cuts into the plateau. The site where I picked up the largest 
number had been buried to a depth of 7 or 8 feet ; only where 
the water after heavy rains had washed great masses of sand and 
gravel away, the arrow-heads with other weapons and implements 
had been exposed. These deeply-buried settlements were doubt- 
less very ancient 1 ." This passage, written by a good observer 
and naturalist, reads like a description of the river-drift finds in 
the Thames and Somme Valleys, and prejudice alone will refuse to 
accept it as proof of quaternary man in America. Here also the 
argument is strengthened by the evident change of climate, which 
at present is far too dry to support the numerous village com- 
munities formerly dotted thickly over the now arid Patagonian 
wastes. Moreno's investigations also establish quaternary man in 

ARGENTINA, BRAZIL. Here the existence of quaternary man 
seems to be established by the researches of Ameghino,' Bur- 
meister, Lund, Moreno and other eminent palaeontologists, who 
have produced not only the works but also the remains of fossil 
man himself, especially from the Brazilian caves and from the 
Pampas beds, which latter answer partly to the pleistocene, 
partly to the pliocene, formations of Europe. The question of 
tertiary man has even been raised by Ameghino, on the ground 
that these beds all belong to the same period, which he refers to 
"pre-glacial," that is, late pliocene times. But Burmeister, whose 
views are confirmed by Soren Hansen, shows clearly that the 
Pampas formations belong to two distinct epochs, the lower alone 
being pre-glacial, the upper quaternary; and as all agree that the 
upper alone contains human remains and traces of human industry, 
the question may be regarded as settled in the same sense that 
it has been settled in Europe and elsewhere tertiary not proven, 
except for a postulated generalized precursor; quaternary proven 
for differentiated Hominidae. The chief localities that have 

1 Idle Days in Patagonia, pp. 37 39. In the same Rio Negro Valley 
Moreno found (1874) at a depth of 13 ft. a skull artificially deformed like those 
of the Bolivian Aymaras (Bull, de la Soc. d'Anthrop. 1880, p. 490). But against 
this supposed widespread practice of cranial deformation a warning note is 
raised by Juan Ignacio de Armas in a paper read before the Havana Anthrop. 
Soc. Nov. 1885, on the so-called deformed Carib crania of Cuba. 


yielded evidence of palaeolithic man are: Lagda Santa district, 
Minas Geraes (Upper S. Francisco basin) ; Sumidouro and other 
limestone caves, explored by Claussen, and especially by Lund 1 , 
who here found the fossil remains of over 30 human beings and 
numerous stone implements associated with an extinct fauna 
answering to that of pleistocene times in Europe; all the skulls 
except one are dolicho- and hypsisteno-cephalic (long, high and 
narrow). Rio Carcarana^ Parana basin (Buenos Ayres Pampas); 
similar remains, including one skull found by Roth under the 
carapace of a Glyptodon near Pontimelo, but of brachycephalic 
type. Some of the fauna present characters like those of the 
tertiary period in Europe; such is the mastodon, which however 
persisted in America long after its extinction elsewhere. Hence 
"there would be nothing strange in the existence in America of a 
mammalian fauna apparently tertiary, but contemporary with 
our quaternary times" (Quatrefages, op. cit. p. 104). Thus here 
again the proof of tertiary man breaks down. On the other 
hand the two different Lagda Santa and Pampas types seem 
to attest the existence of human varieties (the Hominidse) in 
South America in the quaternary period 2 . Samborombon, south- 
east of Buenos Ayres ; human skeleton and megatherium dis- 
covered 1882 by Carles (see p. 34). Santarem district and 
Marajo Island near Para; extensive shell-heaps with skulls of 
same type as the present Tapuyo populations of Amazonia, also 
mounds affecting the forms of alligators and other huge animals 

1 Mtmoires de la Soc. des Antiqitaires du Nord> 1845, and numerous other 
communications. This palaeontologist, who devoted many years to the ex- 
ploration of the hundreds of caves in the Lagoa Santa district, has determined 
as many as 115 species of fossil mammals, including a huge ape, a jaguar twice 
the size of the present Brazilian species, a cabiai as large as a tapir, and a horse 
like that of the eastern hemisphere, but everywhere extinct in America before 
the discovery; all these in close contact with fossil man. 

2 "Les grandes differences que presentent les cranes, les instruments, les 
inscriptions des rochers, prouvent que ces populations appartenaient a des 
souches diverses. Le continent qui se termine en une longue peninsule formait 
comme une sorte de nasse dans laquelle les peuples refoules des contrees du 
nord venaient se pr^ndre les uns apres les autres, et souvent s'entre-exterminer. 
L' Argentine est une vaste necropole de races perdues." Reclus (after Moreno) 
XIX. p. 672. 



(the tribal totems?), resembling the mounds of the Mississippi 
basin 1 . Quixeramobim Valley, Jaguaribe basin, Ceara; skull of 
great age found in a cave, of doubtful Tupi type 2 . Santa Catha- 
rina seaboard ; hundreds of sambaqui (shell-mounds) with human 
remains and rude implements. A skull described by A. Nehring 
( VerhandL, Berlin, Anthrop. Soc. 1895-6, p. 710) shows characters 
like those of Neanderthal, Spy and even Pithec. erectus ; another 
found by Loefgren in a mound 6 miles west of Sao Vicente 
resembled those of the Lagoa Santa caves. The mounds them- 
selves must be of great age, some being overgrown with huge 
forest trees or buried beneath the drift washed down by ancient 
rivers. Many are still over 300 feet wide by 50 high, although 
for over 200 years they have been utilised by the lime-burners of 
Rio Santos and other towns 40 or 50 miles inland 3 . 

MEXICO; fossil human remains found 1884 at foot of the 
Penon de los Banos on the saline plains near the city of Mexico, 
associated with extinct fauna (elephant, horse, c.) beneath a 
lava-stream, indicating a time when the neighbouring Texcoco 
lagoon stood 10 feet higher than its present level and when 
igneous eruptions of remote pre-historic times had not yet taken 
place ; elsewhere numerous palaeoliths also associated with the 
elephant (E. Colombi] point at the presence of man on the 
Anahuac plateau at a time corresponding to the European inter- 
glacial period. 

In the UNITED STATES and CANADA, this period is clearly 
defined between CrolFs ice-ages, the first preceding, the second 
following, the formation of the present Ohio valley ; both in- 
dicated respectively by the normal trend from north-east to south- 
west, and from north to south, of the usual phenomena due to 
the grinding action of the ice-sheets. In a summary such as this 
it would be idle to follow all the vicissitudes of the battle that 

1 F. von Martins, Ethnographic Brasilien's. Many of these vestiges, how- 
ever, are distinctly neolithic or even later, and apart from their associations 
none would suffice to establish the presence of quaternary man in Amazonia. 

2 Lacerda and Peixoto, Coniributf des para o estitdo anthropologico das rafas 

3 "Les sambaqui datent certainement d'ime dpoque reculee...La somme de 
travail que representent ces amas est vraiment prodigieuse" (Reclus, xix. 
P- 359)- 


has not yet been fought out over the presence of palaeolithic man 
in the North American Continent. But speaking generally it may 
be stated that the evidence brought forward even by such eminent 
archaeologists and geologists as Abbott, Putnam, Wilson, Powell, 
Cook, Shaler, and others studying the question on the spot, is 
not yet regarded as conclusively establishing in this region the 
presence of primitive man contemporary of the 
FAiropean pleistocene Hominidce. The proofs 
chiefly relied on consist partly of innumerable sur- Trenton 

J r j gravels, 

face finds from every part of the United States 
and from some Canadian districts (of which presently) and partly 
of numerous palseoliths identical in form with those of the Thames 
and Somme river-drifts, taken from apparently undisturbed glacial 
deposits of corresponding age. Such are those of the Delaware 
valley near Trenton, New Jersey, where Dr C. C. Abbott has 
year after year brought to light from depths of 5 to 20 feet 
scrapers, points and other rude implements of hard argillite, 
one even showing glacial scratchings exactly like those of the 
striated rocks among which it was found. Some were taken 
in situ in the presence of such unimpeachable witnesses as 
Putnam, Shaler, Dawkins, Haynes, of whom the last mentioned 
writes that "speaking from an archaeological stand-point, I do 
not hesitate to declare my firm conviction that the rude argillite 
objects found in the gravels of the Delaware river at Trenton, 
N.J., are true palaeolithic implements 1 ." The Delaware is a 
much larger river (350 miles long) than either the Somme or the 
Thames, and when its banks were assumed to be frequented by 
primitive man its bed stood 50 feet higher than at present. 

Farther inland, evidence has been adduced from a few places, 
such as Claymont (Delaware), Upland (Chester 
County, Pennsylvania) and the glacial gravels of 
Jackson County, Indiana, in all of which districts and other 
palaeoliths are claimed to have been found in situ 
by Dr Hilborne T. Cresson of Philadelphia; the extensive gravel 
beds of the Little Miami near Cincinnati, Ohio, also of glacial 
origin, where specimens were brought forward by Dr C. L. Metz 

1 Quoted by Dr Abbott in Evidences of the Antiquity of Man in Eastern 
North America^ 1888, p. jo. 






at a depth of nearly 30 feet below the surface, so that "we can 
henceforth speak with confidence of inter-glacial man in Ohio" 
(Abbott, ib. p. 6); the drift at Little Falls, Minnesota, where in 
1879 Miss Babbil is stated to have found rudely worked quartzes 
deeply buried beneath the glacial deposits; the Lake Lahotan 
valley, north-western Nevada, where an obsidian spear-head ap- 
parently palaeolithic was found by Prof. MGee. 

At the meeting of the American Association for the Advance- 
ment of Knowledge, Madison, Wisconsin, 1893, all this evidence 
formed the subject of a long discussion, in which it was accepted 
as valid by Prof. G. F. Wright, but impugned by 
Prof. T. C. Chamberlin and others. In closing the v e ?. f 


discussion, Prof. W. J. M c Gee submitted that, Holmes, 
although possible, the existence of man in North others. an 
America even during the last ice invasion of the 
glacial period "had not yet been proved beyond question. 
The supposed evidences of great human antiquity in that country 
had not yet been corroborated by more extended research, but in 
all save one or two cases later research had only served to show 
that the first interpretation was erroneous." This is the view also 
entertained by Mr W. H. Holmes and Prof. Otis T. Mason, two 
most careful observers, both of whom hold that " the finds of 
shaped stones referred to the gravels in place are modern shop 
refuse [rejects, wastrels], involved in the talus deposits in com- 
paratively recent times." After his return from the Chicago 
Exhibition, M. Topinard, reviewing the whole question, expressed 
in V Anthropologie his belief in the high antiquity of man in the 
New World, and alluded to Dr C. C. Abbott as "the Boucher 
de Perthes of America." To this Prof. O. T. Mason 1 replies that 
"it is quite within the limits of possibility that Boucher de Perthes 
may turn out to have been the Dr Abbott of France," meaning 
that his conclusions, since confirmed by overwhelming evidence 
and accepted by Evans, Flower, and even Prestwich and other 
extremely cautious observers, may nevertheless have to be rejected 
as premature. It would appear, on the contrary, that, when not 
merely one section, but the whole field from Fuegia (see above) 

1 American Anthropologist^ Oct. 1893, p. 461. 


to Alaska, is brought under survey, the existence of quaternary 
man in America may be as frankly accepted as it has already been 
in Europe. 

At the meeting of the American Association no reference 

appears to have been made to the famous fossil 
veTas C skun." sku11 reported by Prof. J. D. Whitney as found 

(1886) in the undisturbed auriferous gravels of 
Calaveras County, California, which at once raised the still 
discussed question of " tertiary man " in the New World. Before 
reaching the gravel bed where the skull was said to have been 
found, the shaft sunk by the miners had in downward order 
successively pierced a black lava sheet 40 ft. thick, gravels 3, 
white lava 30, gravels 5, white lava 15, gravels 25, and brown 
lavas 9, or a total depth of nearly 130 feet. As the lavas might 
have accumulated rapidly during periods of great igneous disturb- 
ance in the Sierra Nevada region, everything would depend on 
the age of the gold-bearing gravels, which are assigned by Whitney 
to late tertiary times (pliocene), and by le Conte to "the be- 
ginning of the [last?] glacial epoch." On the strength of this 
and other data Whitney himself concludes generally "that there 
is a large body of evidence, the strength of which it is im- 
possible to deny, which seems to prove that man existed in 
California previous to the cessation of volcanic activity in the 
Sierra Nevada, to the epoch of the greatest extension of the 
glaciers in that region, and to the erosion of the present river canons 
and valleys, at a time when the animal and vegetable creation 
differed entirely from what they now are, and when the topo- 
graphical features of the state were extremely unlike those 
exhibited by the present surface V 1 The question is still sub 
judice ; but should the find prove genuine, it will go some way to 
establish a warm interglacial period of long duration to give 
time for slow movements of migration between the eastern and 
western hemispheres before the second ice-age set in. From his 
studies of the Colombia formation M c Gee infers such an epoch 
for North America, where the relative erosion of running waters 
since the formation of the first (Columbia) and second deposits 

1 Auriferous Gravels of the Sierra Nevada^ p. 288. 


shows that " the interval of mild climate and high level of the 
land between the two epochs of cold was from three to ten times 
as long as the post-glacial period 1 ." 

This would also give time not merely for the appearance of 
palaeolithic man, at a few isolated points, as above, 

f , ,.- . , f , , General dif- 

but for his general diffusion throughout the northern fusion of qua- 
Continent, as some have inferred from the special throughout" 
inquiries made in this direction by Mr Thomas North 
Wilson, Curator of the Department of Prehistoric 
Anthropology in the Smithsonian Institution. From his memoir 
on the subject (Washington 1890), it appears that to a Circular 
(No. 36), issued in 1888, asking for information respecting primi- 
tive man and his works, 209 replies were received, reporting 6,656 
palaeoliths of Chellian and Solutrian (laurel- leaf ) types from 23 
States of the Union and 106 from Canada. Besides these, 
thousands exist in public and private collections, such as those of 
Cambridge, Mass., the New York Natural History Museum, the 
United States National Museum, Washington, the Valentine 
collection, lately presented to the City of Richmond, and the 
Christy, now in the British Museum. 

Some of those reported to the Smithsonian Institution occurred 
in undisturbed deposits, such as those from Warren and Green 
Counties, Ohio ; from Essex County, Mass. ; Bonaparte, Iowa ; 
West Granby, Connecticut (12 ft. below the surface); Lewisburgh, 
Pennsylvania, and elsewhere. But many are from mounds and 
shell-heaps of no great age, while the majority are simply " surface 
finds." A great controversy rages over these, which by many are 
not accepted in evidence, being regarded as wastrels from the 
workshops of neolithic peoples (the present Indians). Flakes and 
chippings of all kinds must be so regarded, unless their age is 
attested by their provenance and associations. But all rudely 
finished implements, say of the Chellian type, are not to be 
rejected merely because found on or near the surface. Often they 
cannot be explained as chips flaked off from the core in the 
process of manufacturing neoliths. They show wear and tear, 
having been used as the best tools palaeolithic man could produce, 

1 Meeting Amcr. Ass. for the Advancement of Science^ 1887. 


and they occur in some places in pockets or caches, as prized 
objects, doubtless rude in a relative sense, but not so to those who 
knew of nothing better. Mr Wilson, one of the first archaeologists 
living, writes (ib. p. 694) : " My experience with these implements 
in the two continents justifies me in identifying those found in 
America as belonging to the same stage of culture to which the 
Chellian implements of France and England belonged, and, con- 
sequently, enables me to call them palaeolithic implements." And 
this must suffice for a subject about which hundreds of papers 
have been written, but on which it would be premature to pro- 
nounce definitely. Hence little has here been attempted beyond 
a fair exposition of the available facts, and of the views advocated 
on both sides. An impartial observer may perhaps be per- 
mitted to add that, if palaeolithic man, as we are told, "is dis- 
credited in the north," he stands in high favour in the south, 
where his existence appears to be placed beyond reasonable 
doubt, at least in Brazil, Patagonia and Fuegia. 

A great antiquity has been claimed for the above-mentioned 

mounds and the earth-works of all kinds strewn 

buiider^not:" 1 " over tne Mississippi basin, and abundant especially 

quaternary. ' m t j le QJ^Q va iw They have been referred to 

Their culture J J 

neolithic, pre- the Tallegwi, an extinct civilised race, ante-dating 

historic and ., ^.TJ-^'I j j 

historic. tne present Indian tribes, and driven out or ex- 

terminated by them. It is confidently asserted 
that between the Mound-builders and the Red Skins " no line of 
connection can be made out," to which it might be replied with 
even greater confidence that "no line of disconnection can be 

Mr W. K. Moorehead 1 , one of the best authorities on this 
subject, recognises two distinct mound-building races, the old long- 
headed, the later round-headed intruders, besides traces of palaeo- 
lithic man near Cincinnati, possibly associated with the mastodon, 
megatherium, mylodon, and huge extinct bears and jaguars, but 
not known to be connected with the mound-builders. The chief 
seat of the long-heads was the Muskingum valley, from Marietta 
upwards to East Ohio, where the mounds, differing in type from 

1 Primitive Man in Ohio^ Boston, 1892. 


those of the round-heads, have yielded pottery, articles of slate, 
hematite, copper bracelets and other ornaments, generally inferior 
to those of the round-heads. These had their chief centre in the 
Madisonville district, at the head of the Ohio river, where have 
been found superior copper, horn, flint, stone, bone and shell 
objects in profusion. Some 24 miles to the north-east are the 
famous earthworks of Fort Ancient, the largest in Ohio, nearly a 
mile long, with over 10 miles of artificial lines. Chillicothe, on 
the Scioto river, is still the centre of the most interesting round- 
head remains, such as the Hopewell group, the Hopeton works, 
the Mound City, and other sites of pre-Shawnee settlements, 
yielding potteries of artistic designs and elaborate workmanship, 
finely wrought flints, copper, and other objects. Moorehead con- 
cludes that none of the mound-building races attained more than 
a high state of savagery, that they were skilled in several arts, but 
excelled in none, that they were not even semi-civilised, much less 
possessors of the "lost civilisation 1 ' with which they have been 
credited. The best authorities l , in fact, now regard them, not as 
a distinct race, but merely as the precursors or ancestors of the 
present aborigines. There is nothing in the mounds that the Red 
Skins could not have executed, and several of these structures 
have been in progress since the discovery. They thus connect 
neolithic and prehistoric with historic times, but do not help in 
any way to bridge over the gap between palaeolithic and neolithic 
man. In the next chapter it will be seen that this problem of the 
continuity of early with later culture everywhere presents itself, 
and nowhere perhaps admits of a complete solution. 

1 Dr Andree (Das Zeichnen del den Naturuolkern, 1887) "reasserts the old 
statement that there is an established difference in artistic capacity between the 
so-called mound-builders and the present Indians, so great that it either shows 
a genetic difference between them, or that the Indians had degenerated in that 
respect. This statement is denied by the Bureau of Ethnology " (Mallery, op. 
cit. p. 738). This may be regarded as decisive, as the Bureau in question, a 
branch of the Smithsonian Institution, Washington, possesses all the materials 
necessary to form an authoritative judgment on the point. See also Twelfth 
Annual Report of Bureau of Ethnology (1894) where the subject is treated 
exhaustively by Mr Cyrus Thomas. 



Marked difference between the Old and New Stone Ages Comparative Table 
of Pakeo- and Neolithic Cultures A Break of Continuity in some regions, 
notably in Britain But not everywhere No universal hiatus possible 
Continuous evolution in the south and south-east Probable duration of 
neolithic times The late palaeolithic era of the West synchronous with 
the early neolithic era of the South-east Great duration of neolithic 
times argued on general considerations The Danish peat-bogs a time 
gauge The Danish kitchen-middens Origin and growth of aquatic 
stations The Swiss Lake-Dwellings The Irish and Scotch crannogs 
Neolithic structures Reducible to two types: The polylith or cell, and 
monolith or block, originating in Burial and Ancestry worship Polylithic 
and monolithic nomenclature Evolution of the Cromlech or Dolmen 
through the Barrow from the Cell Popularly associated with druidical 
rites The Sessi and Stazzone of Malta and Corsica The Nuraghi of 
Sardinia The Talayots of the Balearic Islands The Russian Kurgans 
Silbury Hill The Cell becomes a Family Vault with later develop- 
ments The Menhir, its origin and wide diffusion Its development in 
linear and circular direction The Alignments and Cycloliths (Stone 
Circles) Their origin and purpose explained Erdevcn ; Stonehenge ; 
New Grange; Menec, Carnac district The Irish Round Towers 
Geographical Distribution of the Megaliths Chief Centres: Bahrein 
Islands; Moab; Mauritania; Gaul, Britain, Scandinavia Bearing on the 
question of early migrations Europe re-settled in Neolithic times from 
two quarters Routes indicated by the presence or absence of Mega- 
lithic Structures These wrongly accredited to the Kelts who followed 
the non-megalithic route Astronomic and religious ideas attributed to 
the megalith-builders Prehistoric monuments in the New World General 
Survey Tiahuanaco, culminating glory of American Megrxlithic archi- 
tecture Tiahuanaco Culture an independent local development. 

THANKS to the break of continuity which certainly occurs in 

some places between the old and the new stone 

difference be- a g es > tner ^ is little difficulty in defining the more 

tweentheoid salient features by which these two epochs are 

stone A^es. distinguished. Later, the various grades of human 

culture often merge so imperceptibly one in the 

other, or present such a tangle of survivals and overlappings, 

that it becomes hard at times to say where one begins or the 

other ends. Thus the copper age, which must have preceded the 

CH. VI.] 



bronze, seems, so to say, crowded out almost everywhere in the 
Old World, so that the transition is direct from the neolithic to 


(From various localities in the United States. ) 

the bronze era. Even the bronze seems in some districts fused 
in the iron, as in Belgium, where M. Ch. J. Comhaire is unable 
to determine "the existence of a bronze age in the strict sense, 
but only of a first iron age, that revealed by the Hallstadt necro- 
polis type 1 ." 

1 Bull, de la Soc. cfAnthrop. de Bruxelles^ 1894, p. 18. So at the present 
time we find railways preceding roadways in some newly-settled regions 
(Argentina, the Far West &c.). 




But even where they obviously come into close contact 
(Liguria, Gaul) the men of the palaeolithic age always present the 
sharpest contrast to their neolithic successors. As will be seen in 
the next chapter, the physical types are absolutely distinct, 
except where intermediate forms already point at interminglings. 
All the elements of their respective cultures also differ so pro- 
foundly, as almost to suggest some violent disloca- 
tion or sudden cataclysm, such as those of the 
early geologists, rather than an orderly sequence 
in accordance with the accepted principles of 
organic evolution. The chief differences between 
the two ages may be conveniently tabulated as under : 

Table of 
Palseo- and 

Climate at first warm (inter-glacial), 
then cold (last ice-age) in the present 
temperate zone of the Northern hemi- 
sphere and everywhere in the Alpine 

Fauna: large pachyderms, feline 
and ursine species, hyaena, reindeer, 
horse, elk, glutton, chamois, goat, all 
wild ; some perish with the increasing 
cold, some migrate south, some survive 
by adaptation to the changed environ- 
ment and either withdraw northwards 
with the retreating ice-sheet or take 
refuge in the Alpine regions; no do- 
mestic animals. 

Human types mainly clolichocepha- 
lous, but brachycephalous also in some 
places (South America?). 

Fire, at first known only, later 
partly under control could be pre- 
served when kindled by natural means 1 . 

Food*, at first mainly vegetable, 
then animal also, mostly perhaps eaten 
raw ; obtained by hunting and fishing 


Climate everywhere much as at 
present, though at first (last post- 
glacial period) perhaps cooler. In 
general ice disappears with the ap- 
pearance of neolithic man in the tem- 
perate zone. 

Fauna: mainly as at present, a few 
pachyderms survive here and there 
(mammoth in Siberia); chief wild 
animals wolf, bear, lion, aurochs, 
beaver, fox, deer; domestic animals 
everywhere abundant horse, ox, dog, 
sheep, goat, pig in temperate zone, 
camel in Arabia and Central Asia, 
llama in S. America. 

Human types at first mainly dolicho- 
cephalous in Europe, later mixed and 
diversified as at present everywhere. 

Fire under more complete control 
could be artificially kindled and pre- 
served 1 . 

Food*) vegetable and animal, the 
latter mostly cooked; obtained by 
hunting, fishing, stock-breeding and 

1 " II ne faut pas confondre ces trois choses distinctes: la connaissance du 
feu, 1'usage du feu, la production du feu " (Broca). 

58 It is commonly but wrongly supposed that in the wild state the higher 




Cultivated plants, none. 

Industries limited to the making of 
chipped stone implements of Chellian, 
Solutrian and other types, never ground 
or polished; apparently no pottery, 
but later artistic sentiment developed. 

Momtments, none in the strict sense; 
no houses, graves, or burial. 

Speech, at first perhaps inorganic, 
later involved. 

Religious ideas , none (?). 

Social Groups, the family, later the 


Cultivated plants, numerous, cereals, 
vegetables, fruits. 

Indtistries extended to the making of 
polished stone implements of diverse 
types, spinning, weaving, mining, 
pottery, but little artistic sentiment 
at first. 

Monuments, monolithic, megalithic 
etc. very numerous; houses, barrows, 
graves (burial). 

Speech perhaps everywhere involved 
at first, later organic. 

Religions ideas well developed. 

Social groups, the family, clan and 

Some of these details, such as the comparatively late intro- 
duction of the' art of kindling fire 1 , the true starting 
point in the evolution of civilized man, may perhaps 
be open to doubt. But the table as a whole pre- 
sents a sufficiently accurate picture of the two eras, 
which are here seen to offer the sharpest contrasts 

A break of 

notably in 


apes are exclusively herbivorous. They are certainly also insectivorous and 
carnivorous, eating vermin, eggs, small rodents and birds greedily. "A 1'egard 
des jeunes oiseaux, le gorille et le chimpanze font preuve d'une telle voracite 
qu'ils avalent leur proie sans la deplumer " (L. F. de Pauw, Bull. d. I. Soc. 
d^Anthrop. de BrtixeUes, 1894, p. 140). Hence, when the precursor was driven 
by the increasing cold of the first ice-age from arboreal habits to a nomad life 
on the plains he readily acquired omnivorous tastes. It follows that man, in 
the eolithic stage mainly frugivorous, adapted himself later to a general diet ; 
all physiologists admit that food is largely a question of adjustment to the 
environment, while itself reacting most powerfully on the dentition and 

1 Yet even this may be inferred from the vague reminiscences of the dis- 
covery, which still survived into historic times in the form of the Promethean 
myth. The very names of the two pieces of wood used in one primitive process 
of producing fire are preserved both in Greek and Latin : oropetfs or fox&pa- 
tabula, the stand or under piece; rptiravov = terebra the borer twirled 
between the hands. This "fire-drill," itself an improvement on the still more 
primitive method of the "stick and groove " (Tylor), was in use in connection 
with mystic rites long after it had been superseded for practical purposes by the 
flint and steel, the burning-glass, and other more efficient processes. It thus 




at all points. Hence it is not surprising that a general impres- 
sion should prevail, not of mere sequence, but of an abrupt 
transition without any intermediate stages between the Old and 
New Stone Ages. In some localities, notably in Britain, such 
may have been the case, and Evans aptly remarks that " there 
appears in this country, at all events, to be a complete gap 
between the river-drift and surface-stone [neolithic] periods, so 


(From Bridlington, Yorks.) 

far as any intermediate forms of implements are concerned; here 
at least the race of men who fabricated the latest of the palseo- 

becarae associated with so many superstitious practices, especially in the pro- 
duction of the so-called nodfyr or niedfyr ("needfire ") in Germanic lands, that 
the use of kindling lire by friction (De ignefricato de ligno] was prohibited by 
the Council of Leptines (Hainaut) in 725. Such survivals point to relatively 
recent inventions in neolithic or even prehistoric times. 


lithic implements may have, and in all probability had, dis- 
appeared at an epoch remote from that when the country was 
again occupied by those who not only chipped but polished their 
flint tools, and who were moreover associated with a mammalian 
fauna far nearer resembling that of the present day than that of 
the quaternary times 1 ." 

It has been seen (p. 73) that the same inference is drawn* 
by Prof. Boyd Dawkins, and although questioned by others, this; 


(From the Yorkshire Wolds.") 

certainly seems the most probable view, so fdr as regards Britain, 

where the conditions were peculiar. The few scattered paleolithic 

hunters could scarcely have lived through the last ice-age in a 

contracted region at one time reduced by subsidence to a mere 

cluster of islets, and for long intervals entirely 

severed from the mainland. But elsewhere the everywhere. 

relations were very different, continuous land at all 

times affording a retreat from the advancing ice-sheet, or from 

the glaciers descending from Alpine heights. The southern and 

south-eastern lands (Mediterranean seaboard, Arabia, most of 

Irania and India) not only lay beyond the farthest limits of the 

1 Ancient Stone Implements of Great Britain, p. 612. So great is the 
authority of Sir John Evans on matters of this sort, that his view must be 
accepted until disproved by some direct evidence to the contrary, which is not 
at present forthcoming. 

K. 8 




ice-sheet, but were even favourably affected by glaciation, which 
transformed the temperate to an arctic, and the tropical to a 
temperate zone. Here therefore human culture need never have 
known any break, and if a continuous sequence between old and 
new has not yet been established in these regions, it is only 


(From the Yorkshire Wolds.} 

because they have not yet been everywhere so diligently ex- 
plored as have those north of the Alps. It was seen (p. 73) 
that even in Liguria (Mentone Caves) interminglings seem to 
have taken place 1 ; similar contact may well be suspected in 

1 These, however, are somewhat differently interpreted by Mr A. J. Evans, 
who from the associated ornaments concludes that the three skeletons of the 


France, Belgium, Hungary, and may be assumed for the southern 

and south-eastern lands. In any case no absolute No univcrsal 

or universal hiatus can be imagined without the hiatus 

assumption of one of Cuvier's fresh creations, which possl 

are in themselves a violent and gratuitous assumption, and which 

in science would merely be another name for 

limited knowledge. In this connection it is speci- ^J^ "^* 

ally noteworthy that neolithic man is unanimously the southern 

allowed to have reached Europe from the east or eastern'iands. 

south 1 , most probably from both quarters, and if 

this be so, it follows that those regions were the seat of a 

relatively advanced civilization at the close of the last ice-age in 

the west. This is one of those reasonable inferences which, 

without admitting of direct proof, must yet be accepted in order 

to avoid reckless and incredible assumptions. 

For the whole period, from the close of the last ice-age to the 

Barma Grande Cave, Balzi Rossi Cliffs, are early neolithic although interred in 
unstratified palaeolithic debris. Hence "a race representing the essential 
features of the later population of the polished Stone Age was already settled 
on the Ligurian shores at a time when many of the civilized arts, which have 
hitherto been considered the original possession of neolithic man on his first 
appearance in Europe, were unknown. It will no longer be allowable to say 
that these supposed immigrants from Asia brought with them at their first 
coming certain domestic animals, and had already attained a knowledge of the 
potter's art and of the polishing of stone weapons. And, if this is the case, 
something at least will have been done towards bridging the gap between the 
earlier and the later Stone Age in Europe" (Jour. Anthrop. Inst^ 1893, 
p. 301). 

1 ** Tout le monde reconnatt que celles-ci [les populations ne*olithiques] sont 
venues de loin et ont apporte avec elles des industries jusque-la inconnues sur les 
bords de la V&zere ou dela Lesse, et un etat social nouveau " (De Quatrefages, 
I. 117). Thus even allowing interminglings and contacts at various points, an 
arrest of progress would have still to be admitted for the West. Assuming the 
survival of primitive man into the New Stone Age, it is obvious that in any 
case his culture was interrupted and prevented from continuing its natural 
evolution by the irruption of neolithic man into Europe. One hesitates to 
speak positively on such a difficult question ; but it may be said that all the 
known facts point perhaps at extinction in Britain, and at absorption on the 
mainland. Indeed Mr J. Allen Brown fairly establishes continuity in West 
Europe (your* Anthrop. Inst. 1893, pp. 66 95). 




neolithic and 
ages! 8 On 

present day, a term of over 100,000 years was postu- 

^ ate( ^ at P' 55' ** WaS a * SO Seen tnat ^ tn * S teml 

fully io,ooo years are now required for the strictly 
historic period in Egypt and Mesopotamia. At a 
moderate calculation at least double that number 
of years may be assigned to the prehistoric metal ages inter- 


(Iwerne Minster, Dorset.} 

vening between neolithic and historic times This would leave 
about 70,000 years for the neolithic alone, and a nearer con- 
sideration of the data above tabulated may help to show that this 
is no extravagant estimate. 

From the necessary hypothesis of a neolithic culture syn- 


chronous in the south and south-east with the later 

stages of the palaeolithic era in West and Central palaeolithic 

Europe, it follows that the neolithic era itself, when 
viewed as a whole and not merely in its western chronous with 
developments, must be dated back to palaeolithic Hthic^ra ?n" 
times. In other words, while primitive man was the south- 
still struggling with the mammoth, and fabricating 
chipped implements in Dordogne and Britain, a relatively ad- 
vanced degree of culture had already been developed, say, in 
the Nile and the Euphrates valleys. Consequently the duration 
of this advanced culture is to be measured, not by the first 
appearance of its representatives in the west, but by its first 
beginnings in the east, which may probably have coincided with 
the Madelenian epoch in France. 

How far removed these beginnings are from even the dawn of 
history, may be dimly conjectured by such general Greatdura 
considerations as the following. Not even the tionofneo- 
faintest memory, such as might have been orally JjJJ^JUJf 8 
transmitted in popular myths and folklore, has general con- 

, r .... , siderations. 

been recorded of the origin in time or place of 
any one of the arts characteristic of the New Stone Age. All 
de Candolle's ingenuity has failed to discover, otherwise than by 
inference, the true home of the cereals and other cultivated plants 
already known to neolithic man. No one can even surmise where 
or when weaving, pottery making, and the other early industries 
had their rise. Who can say when man first began to polish his 
stone implements, and fashion them to convenient forms, some of 
which were afterwards perpetuated in bronze and iron with little 
change down to our days? The foundations of the megalithic 
monuments, which yet girdle the globe, are wrapped in im- 
penetrable mystery. Stonehenge and other similar works in 
Britain and Brittany must be of comparatively recent date, for 
their builders had already traversed more than half the eastern 
hemisphere before reaching the Atlantic seaboard. Yet they are 
old enough to have been entirely forgotten by later generations, 
despite the vast labour expended in their erection. Ratio in 
obscure^ says Tacitus (Hist. n. 3), in reference to a rude stone 
pillar representing the Paphian goddess, and the remark may be 


extended to all the works with which neolithic man has covered 
a great part of both hemispheres. 

Some of these remains, however, afford a somewhat more 
_ ^ . u definite idea of the time occupied in their con- 

The Danish r 

peat-bogs a struction. Probably the best time-gauge may 
ime-gauge. ^ e j ia( j f rom a study of the Danish peat-beds, 
rendered famous by the classical labours of Worsaae, Steenstrup 


(Hunmanby^ Yorks.} 

and other distinguished archaeologists. We have already seen 
that palaeolithic man never ranged so far north as Scandinavia. 
Hence the prehistoric remains now found in Denmark go no 
farther back than the polished stone period. Yet so remote is 
that period that since the first arrival of man the climate of the 
country, as indicated by its flora, has undergone not one but 
several successive changes. At the dawn of history the beech 
was, as it still is, the characteristic forest tree, and as it could not 
have sprung suddenly into existence, its general diffusion may 
confidently be dated back to at least 2500 years ago. But the 
peat-bogs, from the lowest depths of which objects of human 
industry have been recovered, disclose three successive layers of 
decayed vegetable matter, showing that before the beech, the land 


was covered with the pedunculated oak, which had displaced the 
sessile oak, successor to the Scotch fir characteristic of a still 
earlier epoch. Allowing from 2000 to 3000 years to each of 
these slow-growing and exceedingly tenacious arborescent species, 
we see that man must have already been in occupation of the 
land at the very least some 10,000 years ago. But to reach 
Denmark he had to traverse the whole of the European mainland 
by whatever route was followed, and such migratory movements, 
always slow, must have taken many successive generations in 
times when most of the land was either forest-clad, or covered 
with vast swampy tracts. Even so recently as the seventh 
century the Gallo-Germanic borderland is still described as "a vast 
region occupied by almost continuous morasses 1 ." In the Danish 
peat-bogs the change of flora roughly coincides with a change of 
culture, the polished neoliths of the fir and early sessile oak 
being replaced by bronze tools which last throughout the upper 
sessile oak and the whole of the pedunculated oak periods, when 
iron comes in apparently with the beech forests, or somewhat 
later. With the neoliths are associated the remains of elk and 
reindeer, but not of the mammoth, which appears to have never 
ranged into Denmark, although in Asia extending far beyond the 
corresponding parallels of latitude. 

In Denmark also were first studied and named the Kjb'kken- 
moddinger, or "Kitchen-middens," which we have ^ ^ . . 

* ' , The Danish 

seen scattered over both hemispheres, but the true Kitchen Mid- 
character of which was determined by Steenstrup, 
Worsaae and Forchhammer. By Lubbock they are referred to 
the early part of the Neolithic age, "when the art of polishing 
flint instruments was known, but before it had reached its greatest 
development*." Surprise has often been expressed that Denmark 
should have proved such an attraction to man at this period. But 
the explanation may lie in the physical and biological conditions 
of a region washed by the warm waters of the Gulf Stream, and 
yielding an abundance of easily captured food. In the middens 

1 "Regio vastis et fere continuis paludibus obsita"; A. G. B. Schayes 
Les Pays-Bas avant et durant la domination romaine, 1838, 1 1. p. 67, quoting 
from Audanus> Lift of St Eloi. 

2 Prehistoric Times % ch. vii. 


are found the shells of the oyster, cockle, mussel and periwinkle, 
as well as the bones of the herring, eel, capercailzie, wild swan, 
duck, great auk (now extinct), stag, roe, wild boar, urus, beaver 
&c., besides the dog probably already trained to the hunt. Here 
consequently were found some of the earliest permanent settle- 
ments, and here was even developed a distinctly local culture, as 
shown by the peculiar and often highly artistic forms of the 
later * stone and bronze implements. This early settlement of 
Scandinavia affords perhaps a clue to the preponderating part 
played by the Norsemen in the course of events in later times. 

Long occupation is indicated both by the great number and 
by the magnitude of the middens, which occur all round the 
shores of Jutland and neighbouring islands, and some of which 
exceed 1000 feet in length, with a breadth of from 100 to 200 and 
a height of 10 feet. A single mound thus contains many tens of 
thousands of cubic yards of refuse. They were certainly of earlier 
formation than the middle peat-beds, for they contain no bronze 
implements, and only a little pottery of coarse type. Since their 
formation the very coast-line has been greatly modified, and the 
Baltic Sea has become so fresh that the oyster, which formerly 
abounded in the archipelago, can no longer live in the surround- 
ing waters. 

To the Danish peat-bogs correspond in point of time the 
lake-dwellings of Switzerland, where analogous 

Origin and ._ ,. . , _ _ ., 

growth of pliysical conditions could not fail to attract some 

stations. f *ke ^ rst neolithic hordes, probably penetrating 

up the Danube valley westwards from Caucasia 
or Asia Minor. Lacustrine or marine settlements form an 
interesting feature in the evolution of human progress, their 
development being intimately dependent on the local conditions 
at certain stages of culture. Communities seated by the shores 
of lakes or shallow inland seas possess obvious advantages over 
tribes confined to the woodlands or the plains. They draw their 
supplies both from land and water, and to their other resources 
are added navigation followed by barter and piracy. But on the 
other hand the wealth thus rapidly accumulated exposes them to 
the attacks of predatory hordes, to guard against which they take 
refuge in their boats. They are thus gradually transformed to a 


floating population, which soon learns to adapt itself to the new 
environment by erecting dwellings on platforms resting on piles 
driven into the mud or sands of a shelving beach. Then, 
when peaceful days and orderly government take the place of 
lawless habits, a return is made to terra firma> and the abandoned 
lacustrine dwellings soon disappear; but the sites remain the safe 
depositories of the multifarious objects of human industry which 
have accumulated beneath the shallow waters during their occu- 

Such is the history, either 'completed or still in progress, of 
the numerous floating habitations which are found in every part 
of the world from the New Guinea coastlands and the estuaries of 
the Borneo rivers to Helvetia and the British Isles, and beyond 
the Atlantic to the aquatic settlements of the Maracaibo Sea, to 
which the surrounding region owes its present name of Venezuela, 
"Little Venice." Such especially is the history 
of the Swiss lake-dwellings, the recent exploration Lake Dweii- 
of which has shown them to be one of the richest ings * 
storehouses of neolithic and prehistoric industries. Antiquaries 
have already explored over two hundred of such 

J r Extend into 

stations, some of which were occupied again and the Bronze 
again, like Hissarlik (Troy), Lachish 1 , and those other Affe ' 
eastern cities, where the vestiges of several distinct civilizations 
are found superimposed one on the other. At Robenhausen, 
south side of Lake Pfaffikon, three such prehistoric occupations 
have been disclosed, each destroyed before the next began, as 
shown by the three sets of piles (100,000 altogether), each pro- 
jecting from 3 to 5 feet higher than the one below. So also 
at Morges, on the north side of Lake Geneva, there were three 
different stations, here, however, not superimposed but standing 
in close proximity within a space of about a third of a mile. 
Nevertheless they were not inhabited simultaneously, but succes- 
sively, as shown by their relics, all stone in the earliest, stone and 
rude bronze hatchets in the next, bronze alone and very fine 

1 The site of this place "was found by Mr F. J. Bliss to contain the 
accumulated remains of as many as eleven cities, which here succeeded each 
other from about 2000 to 400 or 300 B.C.'' (A. H. Keane, Asia, Stanford 
Series, 1895, Vol. i. Ch. iii.). 


bronze in the last, the great prehistoric city of Merges. Even the 
present Morges appears to be some 1200 or 1500 years old; 
yet it never had any record or memory of its predecessor till its 
existence was revealed in 1854 by the subsidence of the lake, due 
to an exceptionally long drought. 

Although the study of the Swiss lake-dwellings dates from the 
year 1854, it should be mentioned that the Irish crannogs had 
already engaged the attention of Sir W. R. Wilde in 1839; in his 
Catalogue of the Museum of the Royal Irish Academy (1857) 
as many as forty-six are recorded as known at that date, and 
many more have since been brought to light. But the crannogs 
"were not, strictly speaking, artificial islands, but duans, small 
islets or shallows of clay or marl in those lakes which are pro- 
bably dry in summer-time, but submerged in winter" (/#.). 
Although true pile dwellings were not unknown, as at Ardmore 
in the South, most of the houses were of the so-called "fascine" 
type, resting not on stakes and platforms but on layers of sticks 
raised above the surface. Hence the connection with the con- 
tinental structures is not obvious, although all alike are referred 
by Dr Robert Munro 1 to the Kelts, the Swiss being also regarded 
by Keller as of Keltic origin. Of the known Irish sites (about 
220) over half (124) occur in Ulster, and nearly all those dis- 
covered in Scotland are centred in the districts nearest to Ulster 
(Ayrshire and Wigtonshire) : they are also of similar fascine type, 
so that here a connection may be established. The Scotch cran- 
nogs were probably constructed by the first immigrants from 
the north of Ireland before they had secured a firm footing in the 
country to which they gave its present name of Scotland. 

It would seem that the settlements on the Swiss lakes were 

far more numerous than those officially recorded. 

Mr Thomas Wilson tells us that he knows many "not 
su C ose<? nly n ted, and where noted as one they really include 

several." He adds: "At Chevroux, Lake Neuchatel, 
I found twelve stations, of which seven belonged to the neolithic 
and five to the bronze age, yet they are noted as only one of 
each. An idea of the extent of these stations may be obtained 
from the fact that they contain from 10,000 to 100,000 piles.... 

1 Jour. Anthrop. Inst. 1886, p. 453. 


At Wallishofen, Lake Zurich, there have been found no less than 
2000 bronze hair-pins, some long with large and beautiful heads, 
which when polished to their original gold colour, must have 
given a gorgeous appearance to the female head-dress of that 
age 1 ." 

Bronze, an alloy of copper and tin in the proportion of about 
9 to i, appears never to have been made in Europe before the 
historic age ; but it was in general use amongst the Egyptians and 
Babylonians many thousand years ago. It certainly was intro- 
duced from the East at a very remote period into Europe, where 
there were numerous prehistoric foundries for recasting worn or 
broken implements. At one of these, near Bologna, some 14,000 
such fractured pieces were found ready to be worked up, when this 
ancient smithery was suddenly closed for ever, by events which 
have passed out of the memory of man as completely as if they 
had taken place in Croll's first ice-age. Bronze was unknown in 
the New World, except in Chimu, where the art of making it 
seems to have been discovered independently. It will be seen 
presently that this fact has a direct bearing on the question of 
the relations between the two hemispheres in remote prehistoric 

For the study of neolithic man, far more important than 
peat-bogs, middens or lake-dwellings, are the multi- 
tudinous megalithic structures which he has strewn 

broadcast over the face of the globe. Despite much reducible to 

two types* 

diversity of form and size, all these structures seem 
reducible to two fundamental types, the polylith or cell, and the 
monolith or block, both primarily associated with burial and 
ancestry-worship, later also with religious rites in the stricter 
sense. As in biology all proceeds from the cell, so in this 
primitive architecture from the corresponding nucleus are evolved 
the various organic structures, which seem to culminate in the 
Egyptian temple (cell), with its obelisks, avenues of sphinxes, 
and other monolithic approaches (block). Clear- 
ness and the exigencies of space will be consulted 

by here grouping in two divisions, according to HtMc nomen- 

. . , - , . clature. 

their affinities to one or other of the two primary 

1 Prehistoric Anthropology, p. 629. 


types, the numerous terms in current use in connection with the 
neolithic monuments : 

(the block). 


Alignments or Avenues. 

Cydoliths or Stone Circles. 


Round Towers. 




(the cell). 

Cromlechs, Dolmens or Quoits. 

Tumuli or Barrows. 
Cairns or Galgals. 
Sessi or Stazzone. 

Here the cell, taken as the starting-point, is essentially a 

sepulchral chamber or tomb, composed primarily of 

th?Cromiech f * ur ^ ve s * x or more megaliths, three or four 

and Dolmen upright or on edge, supporting a horizontal slab, 

through the , . , 111 111 

Barrow or which covers the whole space enclosed, and to 
ttiVceiL 8 fr m which corresponds another horizontal slab, resting 
on the ground as a floor, but not necessarily present. 
Here are deposited the remains of the dead, or else urns con- 
taining their ashes, with or without parting gifts. Then the 
polylith thus constructed is covered with a heap of stones or earth, 
and is called a cairn, tumulus \ galgal, mound or barrow^. But in 
course of time this superstructure may disappear from various 
causes, leaving exposed the original cell, which is then called a 
cromlech or dolmen" ', of which the cistvaen is a mere variety 8 , and 

1 Cairn (from Irish and Welsh earn, rock), a pile or heap of stones, thrown 
together for any commemorative purpose, hence not necessarily containing a 
grave; barrow, from Anglo-Saxon beorh, a shelter, a burial-place (beorgan, to 
shelter), always covers a grave ; galgal, a rough tumulus without a passage for 
secondary burial. 

2 Cromlech (Welsh from, bending, llech, a slab), and dolmen (Kelt, table- 
stone), are practically synonymous terms, indicating any group of uprights 
supporting a flat capstone or table, this table being the original roof of the 
sepulchral chamber. These terms however are not always used with strict 
accuracy, and cromlech especially is often applied to groups of uprights which, 
having no capstone, should properly be regarded as groups of monoliths or 
menhirs, such as are seen in India, Algeria, Brittany and other regions. 

3 Welsh Cistfaen, a chest or box-shaped tomb in a barrow, applied especially 




the quoit a local designation. Such, according to the best 
authorities, would appear to be the genesis of all true cromlechs, 


many of which have been so long exposed that their raison 
has been forgotten. In many parts of Britain, 
Guernsey and elsewhere, they are called "Druids' 
Altars," and are popularly associated with Druidical with Dmidicai 
rites. They are even attributed by some archaeolo- 
gists to the " Kelts," although it would seem more probable that 
the Kelts on their arrival found them ready to hand and utilised 
them for religious purposes. The Kelts certainly did not reach 
Gaul and Britain by the southern route from Syria, through 
Mauritania and Iberia. Hence to them cannot be referred the 

to those receptacles in which were deposited the pots or urns containing the 
cremated remains of the dead. Such cists are still in use amongst the Khasi 
hillmen of Assam, and many appear never to have been covered by a mound. 




numerous structures of identical form found in those 
and h stazzone regions, as well as the Sessi and Stazzone, which are 
of Malta and merely local names for the dolmens of Pantellaria, 

Malta and Corsica. There are no true dolmens in 
Sardinia, where they are replaced by the Nuraghi, abodes not of 

the dead but of the living, though possibly modelled 
TheNuraghi on i on or vanished cromlech prototypes 1 . To the 

and Talayots. e 

same category belong the so-called Talayots, or 
"watch-towers" of the Balearic Islands, which date also from 


prehistoric times, and which are generally supposed to have been 
erected by the same race that built the Sardinian Nuraghi. 

To the British barrows, of which there are two types, the 

older long and the later round-shaped, correspond the Kurgans of 

the Russian steppe lands, and the already described 

The Russian mounds of North America. Both the Kurgans and 

Kurgans. e 

the mounds reach far into the historic period, and 

1 The resemblance of primitive dwellings to the dolmens has often been 
noticed ; but it is reversing the order of sequence to suggest with Miss A. Buck- 
land "that the tombs were reproductions of the houses of the living" (Jour. 
Anthrop. Inst. ix. p. 132). It was surely the other way, for early man, when 
advanced enough to be influenced by religious sentiment, was intensely super- 
stitious, and in his dread especially of his departed ancestry expended far more 
labour on the abodes of the dead than of the living. Innumerable Old Egyptian 
tombs, but not a single Old Egyptian house or even palace, has lasted to our 


the Kurgans were still used as burial-places in the loth and nth 
centuries of the new era. 

In the south-east of Spain, where the gradual transition is so 
clearly seen from the earliest neolithic to the bronze and even 
" silver " epochs, there occurs a type of grave which probably pre- 
ceded the cell or cist itself, just as inhumation certainly preceded 
cremation, which came in with the development of the potter's art. 
"The mode of burial at this [early neolithic] period was by 
inhumation of several bodies in polygonal spaces enclosed by 
stones set in an upright position \ the bodies were interred at a 
slight depth, with knives, arrowheads of flint, and ornaments 
formed of steatite, beads, shells, &c.V Here the superstructure 
seems to be entirely dispensed with, the bodies (as many as fifteen 
have been found together) being interred in the ground, as at 
present, and the sites simply marked by enclosures of upright 
stones ; at least there is nothing to show that the whole was ever 
surmounted by a tumulus of any kind. Later, when cremation 
was practised, the baked clay urns containing the ashes "were 
placed in sepulchral chambers formed of slabs'' (ib. p. 126). 
Here we seem to have the natural evolution of the cell or cist, 
whence the dolmen, as above. 

It thus appears that all graves of the cell type were in principle 
underground, or at least covered, structures 2 . But the super- 
structure was often a laborious and costly affair, 

, . ^ r ... . __ ./ ' SilburyHill. 

such as that of Silbury Hill, near Marlborough, 

Wiltshire, one of the finest in the world, standing on about five 

acres and rising in vertical height 170 and along the 

slope 316 feet. It is obvious that Silbury Hills The cell be- 

* ^ J comes a family 

could not be raised over the grave of every great vault with 
chief, or smaller mounds over those of smaller mems. 6 1013 
people. Hence the same mound had to do duty 
for many generations, and the original cell expanded into the 

1 Henri and Louis Siret, your. Anthrop. Tnst. 1889, p. 124. 

2 Part of the mound is still to be seen, which originally covered the dolmen 
near Corancez, Chartres district, although the huge capstone is no less than 
15 x io feet (A. L. Lewis, Jour. Anthrop. Inst. 1890, p. 68). In the same 
district "there are remains of the tumulus which, no doubt, completely covered 
it" [the dolmen known as "le Berceau "] (ib. p. 70). At another place (Bonne- 




"family vault," developing a more or less complex system of 
lateral chambers, sometimes 30 x 16 and 8 ft. high, with super- 
imposed slabs of corresponding size, some weighing 10, 20 
and even 40 tons. But easy access, with due regard to 
security from the attacks of prehistoric ghouls, attracted by the 
rich offerings deposited with the dead, was a primary necessity. 


We know how this was effected by the pyramid-builders, the very 
form of whose tombs shows that they were merely "petrified 
mounds." So also the neolithic cell had its intricate approaches, 
galleries or corridors 3 or 4 feet wide and sometimes '40 or 50 
long, constructed like the chambers themselves, and blocked by 
cross slabs either at the entrance or the end, or even at both 
entrance and end, of the passage. 

But the dolmen and its gallery being still entirely covered by 
the mound, and not always disposed in the same direction, some 
opening to the north, some to the south, west, east and especially 
south-east, it became desirable or convenient to indicate the 
entrance by some visible landmark. This was effected by setting 

val) both dolmen and encircling menhirs would appear to have been originally 
covered by the barrow, just as " a row of upright stones has been found buried 
in a tumulus in Brittany" (ib. p. 71). Chambered tumuli of the same type 
occur even as far east as Japan (W. Gowland, ib. p. 64). 




up one or more monoliths (the block), generally perhaps two, like 
the two obelisks in front of so many Egyptian temples. Such 
was the menhir, or " tall stone " (Kelt, maen = stone, _. .. . . 

x I ne iviennir, 

Mr - high), apparent germ of all the monuments its origin and 

j j / I ra.- * rp, wide diffusion, 

reduced to our second or monolithic type. Thus 

we see that as the dolmen was originally always concealed from 

view, the menhir on the contrary always stood on the surface, 


sometimes resting on the ground, sometimes sunk a few feet deep, 
sometimes with prepared foundation, according to the size and 
shape of the block. Although no tool marks are now to be 
detected, all appear to have been quarried, the markings being 
blurred or effaced by long weathering. Some, especially in 
Brittany, are of enormous size, those of Penmarck, Cadiou, Mount 


Dol, Plouarzel, Plesidy and Lochmariaquer, being respectively 25, 
28, 31, 36^, 37 and 67^ feet high, and the last mentioned, now 
fallen and broken, weighs no less than 347 tons. Of menhirs 
proper, that is, completely isolated blocks, as many as 739 have 
been enumerated in Brittany alone. They occur also in groups, 
mostly rough-hewn or unhewn, but sometimes inscribed with 
oghams, runes, and other markings, in North and West Europe, 
North Africa, and as far east as the Deccan, the Khasi and Naga 
Hills. Here they are still erected either as votive offerings or as 
monuments to the dead, in association with, or perhaps more 
frequently detached from, the cists or tombs containing the ashes 
of the dead, with which all would seem to have been originally 
connected. " We may trace back the history of the menhirs from 
historic Christian times to non-historic regions, when these rude 
stone pillars... were gradually superseding the earthen tumuli as a 
record of the dead 1 ." In Khasiland, where both vertical and 
horizontal blocks are combined in a single monument, thev 
appear to be in a state of transition, for Mr C. B. Clarke tells 
us that "they are not necessarily placed where the family ashes 
are kept in cists, or near such cists; but they are usually at no 
great distance from the village where the family dwells 2 ." 

But before this divorce took place between menhir and 
chambered tumulus, the combined system acquired 
r a sl "P r i sm g development both in a linear and 

and circular circular direction. When disposed in single, parallel 
or converging rows, the monoliths take the name of 

alignments, and these may be regarded as linear extensions of the 
blocks originally set up at the entrance to the 

mints tnf n " covered passages. But similar blocks are also 

Cycioiiths or found disposed in circular form round the barrows. 

stone -circles. , . , . 

and they are then known as cyclohths or stone 
circles. At present many of the alignments seem to lead 
nowhere, and have consequently remained a puzzle to archaeo- 
logists. Similarly many of the cycloliths seem now to enclose 

1 Fergusson, Rude Stone Monuments, p. 60. 

2 The Stone Monuments of the Khasi Hills, Jour. Anthrop. Inst., 1873, 
p. 486. 


empty space, and have accordingly given rise to much discussion. 
Typical examples are the alignment of Erdeven, 

13 -!l i, I,' j- j ' Erdeven. 

Brittany, where 1120 menhirs are disposed in 13 

converging lines, i mile 536 yards long, and the cyclolith of 

Stonehenge, which is too well known to need 

, T> i- - i JT.UJ Stonehenge. 

description. But when row, circle and chambered 
barrow are studied where all are still found in more or less 
complete structural combination, all seems sufficiently plain. 
Such is the huge domed tumulus of New Grange, 

r -if T-V i i .1 i -Tii New Grange. 

five miles from Drogheda, the largest in Ireland, 
70 feet high, approached by a gallery 63 feet long formed of 
about 22 blocks on either side, the whole enclosed originally by 
a perfect cyclolith, of which about a dozen menhirs are still in 
situ. "The stupendous mound, the circle of enormous blocks 
of stone surrounding it at equal distances, the great masses 
forming the entrance, fill the spectator with wonder at the 
labour necessary to rear so vast a monument. Creeping in 
on hands and knees, we find huge upright blocks from 2 to 
7 feet in height and from 2 to 3 feet 6 inches in breadth, lining 
the entrance passage on both sides, and gradually approaching 
each other [compare the converging rows at Erdeven], until at 
one point farther progress is a little difficult. This point past, the 
passage widens and rises, so that it is soon possible to stand, and 
you find yourself in a chamber nearly circular, with three side 
compartments, two of them containing large stone basins or 
dishes, on or under which I believe the bodies of the entombed 
were placed 1 ." Remove the tumulus and New Grange becomes, 
mutatis mutandis, a Stonehenge on a reduced scale; supply the 
tumulus, and Stonehenge becomes a New Grange on a colossal 
scale. New Grange may thus be likened to those Pacific islets, 
which are encircled by fringing coral reefs ; Stonehenge to those 
Pacific atolls, in which the islets have disappeared by subsidence, 
leaving only a reef- encircled lagoon. The "lagoons," that is, the 
apparently empty spaces, have, when searched, yielded human 
remains, showing their sepulchral character even ages after the 
disappearance of the menhir-encircled barrows. 

1 Miss A. Buckland, Jour. Anthrop. fnst. IX. p. 151. 





So with the alignments, some of which even still terminate in 
cromlechs. Such are those of Menec, near Carnac, Mcncc 
Brittany, a system of 835 menhirs disposed in 1 1 Carnac 
parallel lines 1293 yards long with a cromlech of 
62 menhirs; and Kerlescant, in the same district, with 13 rows 
1000 feet long (258 menhirs in all), and a cromlech square of 
39 menhirs. There are altogether nearly 50 alignments in France, 
and as many as 3500 dolmens of diverse forms, for the most 
part confined to the southern, central and western districts. 

Whether the Irish Round Towers are a transformation of the 
menhir analogous to the Egyptian obelisk, or a de- The Irigh 
velopment analogous to that of the Muhammadan Round 
minaret, as suggested by Mr A. L. Lewis 1 , it is im- ower8 - 
possible to say. Owing to their complete isolation, for they are 
with one or two doubtful exceptions confined exclusively to Ireland, 
they have hitherto remained an unsolved mystery. But they are 
not monoliths, and being cemented they can scarcely be of any 
great antiquity, although some may be pre-Christian, that is, erected 
before the fifth century. Others are certainly recent, although 
these may possibly have been built in Christian times in imitation 
of the older monuments. 

More important, especially in connection with early migra- 
tions, is the subject of the geographical distribution 
of the neolithic monuments. Broadly speaking, and distribution 
excluding mere cairns and earth mounds, which may . f the Me e a - 
be thrown up anywhere, all the stone structures of 
the cell and block types, are mainly confined in Asia to the south 
(Naga, Khasi and Jaintea Hills, the Deccan south of the Vindhya 
Range, Irania, Asia Minor, Moab, Syria, Palestine, Arabia); in 
Africa to Mauritania taken in its widest sense (Tripolitana to 
the Atlantic); in Europe to the south (Crimea, Mediterranean 
islands, Iberia), the west (Gaul, Belgium, and British Isles) and 
the north (Scandinavia). Greece and Italy are excluded because 
the Cyclopean tombs of those regions seem to be of different 
type, and much more recent, being directly traceable to historic 
peoples (Pelasgians, Hellenes, Etruscans). 

1 Jour. Anthrop. fust. IX. p. 144. 


Amongst the chief centres of prehistoric tumuli are: (i) The 
Chief centres Bahrein Islands, Persian Gulf, explored in 1889 by 
Bahrein Mr Theodore Bent, who speaks of "a vast sea of 

sepulchral mounds," comprising " many thousands," 
and "extending over an area of desert for many miles 1 ." Here 
the chambers are two-storied, the lower cemented and reserved 
for the human remains, the upper for parting gifts. The roof is 
formed of flat slabs, the tomb is approached by long passages, 
and the whole mound is encircled by a retaining wall of huge 
stones, exactly as at New Grange. Palm branches were found, 
which had "the flaky appearance of asbestos, showing a great 
change in the climate of this now "desert" region, which Mr 
Bent regards as the probable cradle of the Phoenician race. 

(2) Moab, of which Canon Tristram writes: "The 

Moab. V ' 

three classes of primaeval monuments in Moab, the 
stone circles, dolmens and cairns, exist each in great abundance, 
in three different parts of the country, but never side by side, the 
cairns being found exclusively on the east, on the spurs of the 
Arabian range, the stone circles south of the Callirrhoe, and the 
dolmens north of that valley; one cairn only surrounded by a 
circle of dolmens is found on the north-west.... This fact would 
seem to indicate three neighbouring tribes, coexistent in the pre- 
historic period, each with distinct funeral or religious customs" 
(Land of Moab). But this is not so, for such interminglings occur 
elsewhere, and are to be explained not so much by racial as by 
cultural differences and climatic changes, as above explained. 

(3) Mauritania, a great centre of neolithic culture, 

Mauritania. . ' , , . . . _., , ' 

in some places covered with an incredible multitude 
of every imaginable type of polylith and monolith; described by 
Barth, Broca, and other more recent observers. "These remains 
occur in great variety of form, and in vast numbers, as many as 
10,000, chiefly of the menhir type, having been enumerated in 
the Mejana steppe alone. All kinds of megalithic structures are 
found cromlechs, circles of stones like Stonehenge, cairns, under- 
ground cells excavated in the solid rock, barrows with huge 
capstones, cupped stones, mounds in the form of step pyramids, 
sacrificial altars, even porticos or gateways like those of the Jebel 

1 Proc. R. Geograph. Soc., 1890, p. 13, 


Msid, Tripolitana, formed by two square posts 10 feet high, 
standing on a common pedestal and supporting a huge super- 
imposed block 1 ." (4) Gaul and British Isles, where Gaul 

both types attain their greatest development in the Britain, 

, . , , . . j \ n j- - Scandinavia. 

eastern hemisphere (see above). (5) Scandinavia, 
especially Denmark and parts of Sweden. "They exist in great 
numbers on the west and south coasts [of Sweden], and advance 
nearly to the centre of the land, but they are found almost 
entirely in separate groups, which rarely intermingle; thus the 
chambered tumuli are found massed together between Lakes 
Wener and Wetter, a few being scattered on the south coast, and 
two only on the west, where, as in the south, dolmens without 
galleries, or cromlechs predominate largely. Between these two 
groups, but extending farther to the north, we find a great 
number of cists not covered with tumuli, and a few covered either 
with tumuli or cairns 2 ." Here the "few" explain the "great 
number," representing probably the original condition of all, 
though, as seen, cists are now commonly constructed uncovered 
in the Khasi Hills. 

How is this general geographical distribution to be inter- 
preted, taken especially in connection with the 
above-described special centres of neolithic culture? the question* 
It was assumed (p. 115) that after the last ice-age mirations 
Europe was resettled from two different quarters, 
the east, and the south or south-east. It was also seen that in 
the south and south-east, temperate regions during 
the ice-age, no break need have taken place in the 

normal evolution of human culture from palaeolithic neolithic 

..... T , . . times from 

to neolithic times. Here consequently was the two quarters. 

seat of early neolithic, as later of early historic, 

civilization. But civilization means increase of population, which 

again gives the impulse to migratory movements. 

Hence from these regions, Mauritania especially, dicatedbythe 

must have come the first and the more civilized 

absence of 

stream of migration. Hence also the route fol- megaiithic 
lowed, across the Strait of Gibraltar and along the 

1 A. H. Keane, Africa, 1895, vol. I. p. 73. 
51 Miss Buck land, loc. ?., p. 158. 


west side of Iberia (Portugal) to Gaul and Britain, is everywhere 
strewn with the monuments of these megalithic builders, ap- 
parently a tall dolichocephalic people of non-Aryan speech. 
Later, much later, came the stream of ruder eastern barbarians 
who could build no megalithic structures, and none are to be 
found along the route necessarily followed by them up the 
Danube to Central and West Europe. Here they came into 
collision with their predecessors, with whom they ultimately 
intermingled, driving out some, who perhaps took refuge in 
Denmark and Sweden, whence the megaliths of Scandinavia. 
These eastern hordes would appear to have been a 
wrongly at- smaller race of brachycephalic type, also of non- 

Ar y an speech. Their much later arrival gives time 
follow the non- for the prodigious development of "neolithic archi- 
megahthic tecture," especially in Brittany and the British Isles. 

It is thus seen that this architecture is wrongly 
ascribed to the "Kelts," who certainly arrived by the Danube 
route, or at least from the east, and who before reaching the 
extreme west were long settled in a great part of Central Europe 
(Bohemia, Bavaria, Helvetia, &c.), where they raised no mega- 
lithic structures of any kind 1 . At the same time it is con- 
ceivable, even probable, that after the fusion of the two races in 
the west, the practice of building megalithic structures may have 
been continued for some time, and to that extent the popular 
traditions would be justified. After the universal adoption of the 
language of the conquering Kelts, the earlier element would be 
forgotten except in vague legendary lore, and by later generations 
everything would be attributed to the " Kelts," just as in Mexico 
and Central America everything was in the Aztec traditions 
attributed to the "Toltecs." It is also to be noticed that, coming 
from the east, the Kelts were probably sun-worshippers, and 
may very well have adapted such cycloliths as Abury (Avebury) 
and Stonehenge to the solar cult. Hence their later modifica- 

1 "It is a remarkable fact that no dolmens are found in Central Europe. 
However obscure the origin of the Kelts... there is no possibility of making the 
area of their evolution in space and time to coincide with that of the megalithic 
monuments. In fact the two areas appear to cross each other at right angles " 
(Dr Munro, Jour. Anthrop. Inst. 1890, p. 65). 


tipns, from which Mr A. L. Lewis concludes that "interment was 
at most a secondary object," the "primary object" being "wor- 
ship or sacrifice 1 ." But a survey of the whole field of neolithic 
architecture would seem to show that this is reversing the actual 
sequence, and that burial connected with the first glimmerings of 
the religious sentiment, veneration (fear) of the dead, was the 
true starting point. The connection of these monuments with 
those of Mauritania has been confirmed by M. Ch. Letourneau, 
who finds that many of the carvings on the dolmen des marchands, 
Brittany, are almost identical with those of the so-called "rupes- 
trian inscriptions" of Tunisia and South Algeria 2 . But the Mauri- 
tanian megaliths do not appear to be in any way The Astro 
connected with the advanced religious ideas with nomic notions 
which the builders of the structures in Britain and the^megaiith- 
Brittany are credited by many antiquaries both builders ex- 
English and foreign. M. F. Gaillard, amongst 
others, has endeavoured to show that the alignments of Saint- 
Pierre, and other menhir systems in the Morbihan district, were 
erected "in order to indicate the time of year for celebrating the 
rites and ceremonies in honour of the departed." In support of 
this view he claims that they are disposed in a line either with 
the summer solstice or with the autumn equinox, which he sup- 
poses may still be verified. But if so, the theory itself would 
collapse. All agree that the monuments are some thousand 
years old, for some had already been abandoned and partly over- 
thrown in the time of the Romans. But the position of the 
earth in relation to the sun varies incessantly, although no doubt 
slowly. Hence if the alignments were originally disposed as 
.here assumed, they would be so no longer; and on the other 
hand if they are now so disposed, they could not have been so 
originally. M. de Mortillet also points out that some of the 
systems are coudes, "bent," which again destroys the "solar 
myth 3 ." In general such advanced astronomic and religious 
notions may be conceded with Piazzi Smyth and Norman Lockyer 

1 Jour. Anthrop. hist. 1891, p. 286. 

2 Letourneau and de Mortillet, BnlL d. /. Soc. d'Anthrop. dc Paris, 1893; 
two papers. 

8 Bui. de la Soc. d'Anthrop., June Oct. 1889, p. 424. 


to the Egyptian temple and pyramid-builders, but not with Lewis 
and Gaillard to the rude megalith-builders of Gaul and Britain. 
Great as are the works of prehistoric man in Britannia, Gaul 
and Mauritania, they are rivalled by those of pre- 

m^mem? C in nistoric man in the New World. Reference has 
the New already been made to the barbaric mound-builders 

of the Mississippi basin. South of their somewhat 
formless structures, follow in almost unbroken succession the 
casas grandes of the Pueblo Indians (New Mexico 
anc ^ Arizona); the truncated pyramids and other 
remains of the Toltecs and their Nahua successors 
(Anahuac Tableland); the palace of Mitla (South Mexico) of 
almost classic beauty; the elaborately ornamented temples, 
palaces, "convents," raised by the Mayas of Palenque, Uxmal, 
Chichen-Itza and other cities of Yucatan; the great temples of 
the sun, the causeways, aqueducts and terraced slopes of the 
Peruvian Quichuas. Some of these are prehistoric, while others 
reach well into the historic period. But none can compare in 
magnitude and exquisite finish with the stupendous megalithic 
edifices of doubtful origin, which stand in an almost uninhabit- 
able region near the southern shores of Lake Titicaca on the 
Bolivian plateau, nearly 13,000 feet above sea-level. Although 
often visited and partly described, full justice has only quite 
recently been done to these astounding ruins of 
Tiahuanaco, Tiahuanaco by Herren Stiibel and Uhle, who have 

culminating j^j , i ^ .,1 j 

glory of devoted a sumptuous volume to their description 

and illustration 1 . The monuments, which cover a 
Architecture. large area between the lake and Pumapunga, though 

chiefly centred about the Ak-Kapana hillj . here 
shown to be a natural formation, not an artificial mound, are of 
an absolutely unique character, despite certain general re- 
semblances to the neolithic structures of the eastern hemisphere. 
As shown by the numerous highly polished slabs and blocks lying 
flat on the ground, as if ready for the mason, it is evident that all 
formed part of a general design on a scale rivalling that of the 

1 Die Ruinenstatte von Tiahuanaco im Hochlande des alien Peru, Breslau, 


largest Egyptian temples, but never completed, the works having 
apparently been interrupted by the Inca conquerors about 120 or 
130 years before the arrival of the Spaniards. They must have 
been in progress for some generations before that time, for the 
blocks, some weighing from 100 to 150 tons, had been conveyed 
with primitive appliances from distances of many miles over 
rugged ground, up steep inclines, and in some cases across several 
inlets of Lake Titicaca. A number of the blocks are disposed as 
uprights like those of Stonehenge, with shoulders for the re- 
ception of horizontal connecting beams, but far better dressed 
and mortised. Others form doorways hewn in a single piece, one 
of which at Ak-Kapana is the crowning triumph of the primitive 
American architecture ! . This marvellous monolith, weighing over 
12 tons, is richly carved on one face with symbolic devices and 
the image of Viracocha, tutelar deity of the Bolivian Aymaras, 
overthrown by the Quechua worshippers of the rival Peruvian 
sun-god. When the sway of the Incas was spread over the 
whole of the middle Andean plateau, there was no longer 
room for two independent and hostile religious centres Pacca- 
ritambo and Tiahuanaco, the "Gerizim and Ebal" of the New 
World; hence the political subjection of the Aymaras to the 
Quechuas was followed by the inevitable suppression of the 
Viracocha cult, and the arrest of the Tiahuanaco works by the 
Incas, shortly before the suppression of the Incas themselves 
by the Conquistadores. Such was the origin and end of this 
splendid Aymara culture, in which the transition is clearly seen 
from the rude and inorganic buildings of neolithic to the true 
"megalithic architecture" of historic times. Not that the Tia- 
huanaco works are to be connected with those of Tiahuanaco 
the eastern hemisphere, or even traced with Angard, Culture an 

1 independent 

Clements Markham, Middendorff and others to local develop- 
Toltec, Maya or Inca sources. Despite the mis- ment * 
leading statements of Garcilaso de la Vega, blindly followed 
because of his Inca descent by most archaeologists, Stiibel and 
Uhle make it clear that the Incas were not the founders but 

1 "Seine Bedeutung iiberragt...alles was bis jetzt in Peru aufgefunden 
worden ist. Es zahlt unter den merkwiirdigsten und interessantesten Resten 
des vorcolumbischen Amerika " (Ruinenstdtte^ Text, p. 20). 


the destroyers of Tiahuanaco, and also that this culture had 
its origin neither amongst the Mexican Toltecs, nor the Mayas 
of Yucatan, but is to be regarded as an independent local 
development amongst the Bolivian Aymaras, elder brothers of 
the Peruvian Quechuas. But "an independent evolution of 
different social systems in different environments seems to 
be a view still beyond the grasp of a certain school of 
ethnologists and antiquaries, who run to the ends of the earth 
seeking 'affinities' and ' origins' and * influences' where none 
exist, and who 'affiliate' two cults or two peoples, no matter 
how many continents and oceans may intervene, if only both 
worship the same sun and moon, forgetting that after all there 
is but one sun and one moon for people on this planet to choose 
from 1 ." If all peoples, as will be seen in the next chapter, not 
only come of one stock but have, relatively speaking, diverged 
but little from their pleistocene precursors, is it surprising that 
resemblances and parallelisms of all kinds should occur in their 
independent later evolution? 

Almost the concluding words of Dr Robert Munro's address 
to the Anthropological Section of the British Association, 1893, 
were that "man's immense antiquity is now accepted by a vast 
majority of the most thoughtful men." Possibly some of the less 
thoughtful may also accept the same conclusion from the con- 
siderations set forth in the foregoing pages. 

1 A. II. Keane, Academy, July 8, 1893, p. 37* 



Specific or Varietal unity decided by extent of divergence between past and 
present races Species and Variety The Physiological test : inter-racial 
fertility The Canidse, Equidse and Hominidse The Palaeolithic races 
Their remains : Trinil : Homo Neanderthalensis ; LaNaulette; LaDenise; 
Spy ; Kent ; Podbaba ; Predmost ; Marcilly ; Mentone ; Olmo ; Eguis- 
heim ; Laugerie ; Palseolithic races exclusively long-headed Neolithic 
races at first also long-headed, then mixed, and later exclusively round- 
headed in some places But all intermingled Fertile miscegenation 
established for prehistoric times In the historic period mixture the rule, 
racial purity the exception The Mestizos of Latin America The Paulistas, 
Franco-Canadians, and Dano-Eskimo The United States Indians and 
half-breeds Eugenesis established for the New World, and for Africa : 
The Griquas, Abyssinians, Sudanese, and West African Negroes Mixed 
races in Asia, Malaysia, and Polynesia The Pitcairn Islanders The physio- 
logical test conclusive against the PolygemstsThe anatomical test The 
Polygenist linguistic argument: Independent stock races inferred from 
independent stock languages P'allacy of this argument Specific Unity 
unaffected by the existence of Stock Languages which are to be other- 
wise explained The Monogenist view established and confirmed by the 
universal diffusion of articulate speech Psychic argument The question 
summed up by Blumenbach. 

IN the address referred to at the end of the last chapter it is 
also stated that u all the osseous remains of man specific or 
which have hitherto been collected and examined varietal unity 
point to the fact that, during the larger portion of extent of y 

the quaternary period, if not, indeed, from its very 
commencement, he had already acquired his human and present 
characteristics. This generalization at once throws races> 
us back to the tertiary period in our search for man's early 
appearance in Europe." It was seen (p. 32) that a tertiary 
generalized form has been fairly well established by Sergi. The 
" human characteristics" of the quaternary "osseous remains" 
have now to be considered, with a view to determining the 
extent of their divergence from each other, as well as the extent 


of the divergence of the living primary divisions both from these 
quaternary prototypes, and from each other. It is obvious that 
on the extent of these divergences depends the question of 
man's specific or generic unity. 

It is not always easy to draw the line between species and 

mere variety, more especially as to neither of these 

terms is any longer attached the idea of finality. 

But, speaking broadly, species may be said to 
possess a large measure of stability, whereas variety is essentially 
unstable, holding an intermediate or transitional position between 
species and species. Variety is species im Werdcn, as the 
Germans would say; that is, a form breaking away from a specific 
type, and tending to become itself a new specific type. When, 
therefore, it is here said that the Hominidae, past (quaternary) and 
present (living), are varieties, not species, all that is intended is 
that the forms diverging from a common precursor are still 
relatively speaking unstable, not having yet reached that stage 
which constitutes true species. 

Here we have two assumptions, both strenuously denied by 
many ethnologists, first, that the Hominidae descend from a single 
precursor, secondly, that their differences are comparatively 
slight, or not sufficiently pronounced to be regarded as specific. 
But both points may, so to say, be determined by one con- 
sideration. It is mainly a question of physiology, and all physio- 
logists are now of accord in accepting fertility as the ultimate 
test of varietal and specific difference. Species and sub-species, 

varieties and sub-varieties, may, as pointed out by 
" Darwin, "blend into each other by an insensible 

l series," giving "the idea of an actual passage 1 ." 
But, however imperceptible the transitions, they 

are continuous only so long as fertility persists; where fertility is 
arrested true species is established. The various 

and^qutdae* breeds of dogs differ far more from one another in 
respect of form, colour and texture of the hair and 

relative size, than do the Equidoe from one another; compare 

on the one hand the skye or the toy terrier with the blood- 

1 Origin of Species > 2nd ed., p. 41. 


hound or bull-dog ; on the other the ordinary horse with the ass 
or even the zebra. Yet all dogs are grouped in a single species 
of the Canidae, being held to be mere varieties, because where 
pairing is possible they are permanently fertile among them- 
selves. But the Equidae form contra so many distinct species, 
despite their much closer general resemblance, because the cross 
is a mule. Some zoologists have even spoken of a specific 
identity of dog and wolf; but nobody denies the specific differ- 
ence of horse and ass. 

Applying this severest of tests to the Hominidae, it is found 
that none breed mules, but that all have been per- 
manently fertile amongst themselves since quater- Homi ~ 

nary times; consequently that they form varieties, 
not species, and are sprung from a single precursor. As regards 
the past, that is, the palaeolithic, neolithic, and prehistoric eras, 
the point is established for all who accept the general con- 
clusions of the leading French and English anthropologists re- 
garding the "universal miscegenation" of primitive man with 
later immigrants in Europe, a miscegenation proved by the 
persistence of pleistocene characters down to the present time 
(see pp. 34-5). For others, who may perhaps not unreasonably 
feel somewhat sceptical regarding this persistence of palaeolithic 
characters, a nearer study of primitive man himself may supply a 
stronger argument for the specific unity of mankind. 

Fully authenticated remains of palaeo- or even of early neolithic 
man are not numerous, and those hitherto brought 
to light are mainly confined to restricted areas 
Europe (especially France and North Italy), and 
South America (Brazil, Argentina). The reason is obvious. 
Interment appears not to have been practised by the river-drift 
hunters and other even earlier generations. Hence none of their 
osseous remains could survive except the few that might have 
been preserved in caves and rock-shelters. It is accordingly in 
such "hermetically sealed receptacles" that have been found the 
skulls, and in still rarer instances the imperfect skeletons now 
available for the study of primitive man. Subjoined is a tabu- 
lated summary of results. The consideration of these might 
indeed be dispensed with if Virchow's statement that "scientific 




anthropology begins with living races," could be accepted. But 
when Virchow 1 himself tells us that "the first step in the con- 
struction of the doctrine of transformism will be the explanation 
of the way the human races have been formed," it is evident that 
the oldest known precursors of the present races cannot be over- 
looked. The truth should now be frankly stated, that, as in the 
case of Cuvier and Owen, Prof. Virchow's vast knowledge and 
range of thought have been somewhat neutralized by his excessive 

TRINIL, left bank river Bengawan (Solo), Java; roof of skull, 

an upper molar, and a femur found (1891 4) by Dr Eugene 

Their Dubois in pleistocene (?) bed 12 to 15 metres 

Remains: below the surface, showing characters intermediate 

between gorilla and Neanderthal, but distinctly 

human; low depressed cranial arch; index 70; capacity iooo(?); 


( Upper surface of skull.} 

very narrow frontal region ; highly developed superciliary arches ; 
"the lowest human cranium yet described, very nearly as much 
below the Neanderthal as this is below the normal European"; 

1 Popular Scienct Monthly (quoted), Jan. 1893, p. 373. 


femur quite human, 455 mm. long, showing height 1654 mm., 
that of an average Frenchman, but found 12 or 15 metres from 
the skull, hence may not belong to the same individual; same 
remark applies to the tooth, which is very large, but more human 
than simian. For these remains Dubois forms a new family 


a. Ordinary Irish skull ; b. Spy cranium ; c. Neanderthal 
cranium; d. Pithecanthropus; e. Gorilla. 

(Pithecanthropus erectus, cine Uebergangsform aus Java, Batavia, 
1894); but they cannot represent a transition between man and 
any of the existing Anthropoids, Pithecanthropus standing in the 
direct human line of divergence in the genealogical tree, although 
considerably lower down than any human form yet discovered 
(Dr D. J. Cunningham, paper read at meeting R. Dublin Soc. 
reported in Nature, Feb. 28, 1895, p. 428). 

NEANDERTHAL (see p. 33); a brain-cap, two femora, two 
humeri and some other fragments, now in the Homo 
Fuhlrott Collection, Elberfeld; normal character Neander- 
established by Broca against Virchow's pathological thalensis - 
theory; remarkable for its flat retreating curve; dolichocephalic 
(index 73.76); frontal sutures closed, occipital more or less free; 
enormous superciliary ridges; the most ape-like skull next to 
Pithecanthropus erectus; in the vicinity were found remains of 
rhinoceros, cave bear and hyaena; Chellean epoch (Fuhlrott, 
Huxley, Broca). 


LA NAULETTE, near Dinant, Belgium; an imperfect lower 

jaw found by Edouard Dupont in a large cave on 

the left bank of the Lesse, which joins the right 

bank of the Meuse above Dinant ; was associated with remains 

of mammoth, rhinoceros, reindeer; now in the Brussels Natural 

Hist. Museum ; simian characters very pronounced in the extreme 

prognathism and alveolar process (teeth themselves lost), canine 

very strong, large molars increasing in size backwards (Dupont). 

LA DENISE, Espaly-Saint-Marcel commune, Upper Loire; 

two depressed and retreating frontal bones from an 

La emse. argillaceous limonite under the muddy bed of an 

extinct volcano; now in the Pichot Collection and Muse'e du 

Puy; (neighbouring beds have yielded remains of cave bear and 

hyaena, mammoth, large hippopotamus, rhinoceros tichorhinus) 

glabella of one very prominent, recalling the Neanderthal ; that of 

the other also prominent and separated from the retreating 

frontal bone by a deep depression; superciliary ridge of both large 

and thick (Aymard, Sauvage, Hamy). 

BRUX, near Prague, Bohemia; a brain-cap and other bones 
from a quaternary sandpit; now in the Vienna Anthropological 
Society's Collection; frontal region and flat elongated parietals 
like those of Neanderthal and Eguisheim, but superciliary bosses 
larger than the latter (Woldrich, Rokitanski). 

SPY, Betche aux Roches cavern, left bank Orneau R., Namur 
district, Belgium ; two nearly perfect skeletons (man 
and woman) found 1886 by Maximin Lohest and 


Marcel de Puydt at a depth of 16 feet, with numerous im- 


plements of the Moustierian type ; now in the Lohest Collection, 
Liege; enormous superciliary ridges and glabella; retreating 
frontal region ; extremely thick cranial wall ; massive mandibular 
ramus with rudimentary chin ; large posterior molars ; divergent 
curvature of bones of the fore-arm ; tibia shorter than in any other 
known race, and stouter than in most; tibia and femur so 
articulated that to maintain equilibrium head and body must 
have been thrown forward, as seen in the large apes. These and 
other characters "place the man of Spy in the lowest category... 
the dentition is inferior to that of the neolithic man in France... 
approximates near to the apes, although there is still, to use the 
language of Fraipont and Lohest, an abyss between the man of 
Spy and the highest ape" (Cope, op. rit. p. 334); associated fauna, 
woolly rhinoceros, elephas primigenius, cave bear and hyaena, &c., 
five extinct, four existing species (Fraipont, Lohest, Cope). 

nearly perfect skeleton found by Mr R. Elliott and 
Mr Matthew Heys in situ at a depth of 8 feet in 
the Pleistocene high-level gravels about 90 feet above the Thames, 
with numerous palaeolithic implements and remains of extinct 
mammals close by; skull hyperdolichocephalic, extremely long, 
narrow and much depressed, with height and breadth indexes 67 
and 64; glabella and brow-ridges prominent; forehead somewhat 
receding; all chief sutures obliterated; three lower molars and 
two premolars in place ; last lower molar, which in Neolithic 
skulls is smaller, is in this specimen as large, if not larger than 
the first ; height about 5ft. i in. ; altogether most nearly related 
to the Neanderthal, Spy and Naulette types (Dr Garson); "is the 
best authenticated record of the occurrence of human remains 
in the higher river-drift that has 'yet been brought forward in 
England" (J. Allen Brown). From the anatomical characters 
Prof. Sollas thinks it highly probable that the remains were in a 
natural position and of same age as the gravels, and not merely 
interred in them at a later (Neolithic) period, as suggested by Sir 
J. Evans and Prof. Boyd Dawkins (E. T. Newton, Meeting Geolog. 
Soc. May 22, 1895). 

PODBABA, near Prague, fragment of skull found 1883 in 
undisturbed brick-clay 13 ft. thick near remains of mammoth, 

IO 2 


rhinoceros tichorhinus, reindeer, &c., at same level; in the large 
superciliary ridges and depressed frontal region 
approaches the Neanderthal type 1 (Fritsch, Kerck- 
hoffs, de Mortillet, LHomme, 1884, p. 528). 
PREDMOST; see p. 34. 

MARCILLY-SUR-EURE, Evreux district; part of skull also of 
Neanderthal type, exposed by a railway cutting at 
a depth of 22 feet, now in Dore'-Delente Collection, 
Dreux ; frontal separated by deep depression from the super- 
ciliary ridges; frontal suture completely ossified (de Mortillet, 
LHomme, 1884, p. 48). 

ARCY-SUR-EURE, Yonne, Grotte des Fe'es ; lower jaw found in 
contact with rhinoceros teeth and Moustierian implements; some- 
what modified Naulette type, prognathism less pronounced (Broca, 
de Quatrefages, and Hamy). 

BALZI Rossi CAVE, Mentone, Liguria, see pp. 


73> I][ 4- 

OLMO, near Arezzo, Tuscany; skull exposed in 1863 by a rail- 
way cutting, 50 feet below the surface in blue lacustrine marl bed, 
with remains of elephant and a Moustierian implement ; now in 
the Florence Geological Museum ; nearly as dolichocephalous as 
the Neanderthal, but superciliary ridges flat and frontal high ; but 
"judgment must be suspended on this find, surrounded as it is 
by so much doubt" (Salmon, op. 'tit. p. 17). 

EGUISHEIM, near Colmar; part of skull found in 1865 associated 

with elephas primigenius, now in the local Museum : 

Eguisheim. i i 6 > > 

prominent superciliary ridges ; frontal region broad 

but retreating; sutures very simple and nearly effaced; marked 

dolichocephaly (Faudel, de Mortillet). 

LAUGERIE-BASSE, Tayac district, Dordogne; one skeleton 
(male), two skulls (female); thick parietals, cranial 
capacity above the modern average in the male 
and in one female skull, but in the other female 

very low, about 1,100 cc.; all dolichocephalic (Lartet, Christy, 


1 "The bone has nearly the same appearance as those of the diluvial 
mammals found in the same clay, commonly considered fossil " (Dr Anthon 
Fritsch, Science, June 27, 1884, p. 786, where illustrations are given). 


The foregoing all belong to the various palaeolithic epochs, 
and while all without exception are dolichocephalic 

/ j r i_ / ,. \ i_ j- / .1 Palaeolithic 

(index ranging from about 70 to 75); the distinctly races ex- 

low characters show progressive modifications in 
the direction of the higher neolithic and modern 
types. But the general assumption that brachycephaly appears at 
once with the neolithic age is certainly a mistake. On the con- 
trary, when all the evidence is sifted and correlated, it will pro- 
bably be made manifest that dolichocephaly even of a pronounced 
type persisted far into neolithic times, and that it was only very 
gradually first modified and then replaced in some regions by 

In the last Chapter it was shown that Europe was re-settled 
from two different quarters, by megalithic builders 
from the south, and from the east by rude hordes races at first 
who nowhere raised stone structures, and it was SI 180 . 1 *? 8 " 

' headed. 

suggested that the former were a dolichocephalic 
(long-headed), the later a brachycephalic (round-headed) people, 
who arrived in the west at a much later period. If so, and it will 
be seen that all the facts point in this direction, the persistence 
of a long-headed type will be at once explained. In Britain, 
where there was, so to say, a tabula rasa owing to the general 
disappearance if not the actual extinction of palaeolithic man, 
English archaeologists are unanimous in holding that the round- 
headed builders of the round barrows were preceded by the 
long-headed builders of the long barrows. Consequently in this 
region dolichocephaly is established for early neolithic times. 

So in France, Belgium, Italy and elsewhere the later cave- 
dwellers and the early dolmen-builders appear to have been all 
first of long, then of medium, and lastly in some places of ex- 
clusively round-headed type. Thus the " Cro-Magnon race," as 
it is called by French anthropologists from the numerous remains 
found by Lartet, Christy and others in the cave of that name at 
Eyzies, Tayac district, Pe'rigord (Dordogne), shows a mean cephalic 
index 73*34 (Broca), hence was distinctly long-headed. This 
race, however, although most probably early neolithic, is regarded 
by some as late palaeolithic. But there can be no doubt about 
the neolithic age of the remains from the dolmen of Maintenon^ 


Eure-et-Loir (Index 73*54, Broca); from the Matarelle cave, 
Aveyron (index 73*62, Durand de Gros); from the Sorgues, 
also Aveyron (73*70, de Gros); from the station on Lake Ladoga, 
(ten skulls, mean 73*64, Bogdanov); from the Soutane Cave, 
Navares-de-Aguso, Spain (ten, mean 73*96, Verneau); from the 
JBaumes-Chaudes Cave, Lozere (thirty-five, mean 74*06, Broca). 
Even the numerous skulls from the Caverne de VHomme-Mort 
Then mixed ("Dead Man's Cave"), Saint-Pierre-de-Tripiez, 
and later Lozere, are all long except two intermediate (mesa- 

round headed ticephalic), seventeen ranging from 68*21 to 76*66, 

insome places. and ^ Qmy rising ^ ^.^ an( j ^^ ( Broca)> After 

this round heads begin to appear, as in the dolmens of Lozere 
(18 long, 6 round, one medium, Broca); and in the sepulchral 
chambers of the Petit-Morin Valley, Marne (22 long, 10 medium, 
1 2 round, Broca). Then the round grow more numerous, and at 
last outnumber the long, as at Orrouy, Oise, (7 round, 4 medium, 
5 long), and the dolmen of PEtang-la-Ville, Seine-et-Oise (3 round, 
no long or mean, Chudzinski). "Towards the close of the neo- 
lithic age in France, the round and medium types become eight 
or ten times more numerous .than the long in certain regions 1 " 

Similar evidence is yielded by the neolithic caves of Finale 

B 11 ai'ke (Ligurian Coast), and other parts of Italy (Dr G. 

everywhere Nicolucci); by those of Hallstatt, Austria, and by 

intermingled. Furfooz and others in Belgium (Dupont), as well as 

by the British round barrows, where all types are associated in 

varying proportions. In England the round heads seem to come 

in with the metal age, as shown by the contents of their barrows, 

and it is evident that here, as on the mainland, the two types 

were gradually merged in a mixed population, 

Fertile mis- which with later superadded elements '(Kelts, Teu- 

cegenation * x 

established for tons &c.), persists to the present time. Permanent 

miscegenation, that is, mixed races capable of trans- 
mitting their kind, is consequently established for 
the prehistoric populations of Europe. Thus, however they may 
have differed from each other in outward form, the primitive 

1 Ph. Salmon, op. dt. p. 39. To this palaeontologist is due the credit of 
having correlated these important data. 


peoples of this region are shown by the physiological test to 
have been varieties of a single species descended from a pliocene 
precursor, and the diagram, p. 38, is justified. 

Throughout the historic period the same phenomenon of 
fertile miscegenation is everywhere presented, to 

& , J . r - , , - In the his- 

such an extent that amongst the present inhabit- toric period 
ants of the globe the rule is mixture, the exception ^ie* U ro C iai the 
racial purity. When comparative anatomists, such purity the 

, ~ . exception. 

as Broca, Flower, or Garson, cast about for speci- 
mens of absolutely pure types, they have to explore such secluded 
upland valleys as those of Savoy or Auvergne, or else extend their 
enquiries to remote insular groups, such as the Andamans, Fiji, 
Tasmania, or Fuegia, and even then they are not always sure. 
From large ethnical groups Malays, Mongols, Germans, Sudan- 
ese little is to be gleaned except averages, mostly of doubtful 
value. But averages mean transitions of all kinds, and transitions 
could result only from extensive interminglings, which again could 
take place only between varieties. 

This general inference will be confirmed by a closer survey of 
the whole field. For this purpose mankind may ^ 

. r J The Mestizos 

be divided into two sections, the older groups of Latin 
whose mixed character can only be indirectly in- Amenca - 
ferred from the foregoing considerations, and the more recent 
groups, whose mixed character can be proved by direct evidence. 
Of the latter by far the most important are the present inhabitants 
of Central and South America, the immense majority of whom 
are confessedly mixed peoples Lusitano-Americans with a con- 
siderable strain of Negro blood in Brazil, Hispano-Americans 
elsewhere. "Whatever be the pretensions of certain sections of 
the community, there can scarcely exist in Latin America any 
really pure race, for the first European immigrants from Mexico 
to Chili nearly all married native women, and since then twelve 
generations have followed, diversely modified by unions between 
every shade of half-breeds. The American populations, which in 
virtue of these unions belong at once to both races may be esti- 
mated at about thirty millions altogether 1 ." 

1 Reclus, English ed., xv. p. 53. 


But it is pretended that these mestizos are not a stable race, 
and would disappear or revert to one of the primitive types, but for 
the constant infusion of fresh blood from Europe. Scarcely any 
immigration 1 however is directed towards Mexico, the Central 
American States, Columbia, Venezuela, Peru or Bolivia; yet in 
all these regions there is a steady increase of population despite 
epidemics, and physical and political convulsions. Thus : 

Mexico (1874) 9,343> ( l8 9 r ) 11,643,000. 

Salvador (1886) 651,000; (1892) 780,000. 

Columbia (1810) 1,000,000; (1892) 4,200,000. 

Venezuela ,, 800,000; 2,323,000. 

Peru 1,100,000; 3,000,000. 

Bolivia 800,000; 2,350,000. 

In Brazil the famous "Paulistas" (so called from the province 

The of Sao Paulo), a cross between the first Portuguese 

Paulistas. immigrants and the aborigines, have always been 

the most vigorous and enterprising section of the community. 

Mainly to them is due the extension of the Portuguese domain from 

the Atlantic seabord to the eastern slopes of the Cordilleras. In 

Canada the French element was probably saved from extinction 

by its alliance with the surrounding Algonquian tribes, and the 

The Franco stur( ty Franco-Canadian boatmen and voyageurs 

Canadians and (trappers and traders) yield to none in energy and 

s imo. ph vs i ca i vitality. In Greenland also the Dano- 

Eskimo half-breeds are not only a thoroughly constituted race, 

but they also show qualities in some respects superior to those of 

either of the original stocks. 

In North America proper, Dr Franz Boas, who has made 
a special study of "the anthropology of the half- 
Cd breeds," declares that "the present generation of 

Indians and Indians is mixed to a considerable extent with 

rialt-Dreeus. . 

whites and negroes, so much so that in certain 
regions it is impossible to find a full-blood individual. Thus the 

1 Nearly everywhere immigration is almost nil, or balanced by the 
emigration, the only exception being Peru, which receives a small but steady 
supply of coolies, chiefly Chinese; these number at present (1895) altogether 
about 50,000, but most of them return to China at the end of the contract time. 


numerous tribes of the Iroquois, Cherokees, Chickasaws and Choc- 
taws contain very few full-blood individuals, if any 1 ." It would 
further appear that while the Indians, as a whole, are decreasing 
owing to various social, political and other extraneous causes, the 
decrease is mainly confined to the unmixed element. "Indian 
women of more than forty years have as an average, approximately, 
six children, while half-breed women have on an average from 
seven to eight... The smaller numbers of children are very much 
more frequent among the Indians than among the half-breeds." 
In a word "we find the rather unexpected result that the fertility 
among half-breed women is considerably larger than among full- 
blood women (ib, p. 39). 

Thus what Broca calls eugenesis, that is, indefinitely fertile 
miscegenation, is proved for the whole of the New 
World. It is also proved for South Africa by the 

r r *t. i ^ j established for 

persistence for over 200 years of the "bastaards, the New 

that is, the Hottentot-Dutch half-castes known as 

Griqnas, who form flourishing communities in TheGriquas. 

Griqualand West and East ; and also by the Negro- 

Hottentot half-castes known as Gonaquas, "Borderers," scattered 

in small groups over the eastern provinces of Cape Colony. 

Farther north the Gallas, Somali and Abyssinians 

of North-East Africa are certainly a blend of the Somali, 

Negro and Hamite on the one hand, and of the 

*-* oouaanese, 

Negro, Hamite and Semite on the other. Most of West African 
the Soudanese populations also, Mabas, Baghirmi, e roes - 
Dasas, Kanuri, Kansas, Songhrays, "Toucouleurs," Fulahs are 
not negroes, but negroid mixtures of Hamites and aborigines all 
along the borderlands between the Berber and Black domains. 
Some of these, notably the Hausas, are greatly superior in many 
respects to both of the primitive elements. From a careful study 
of the West African negroes (Senegal to Angola), J. Deniker 
and L. Laloy conclude generally that they also are a mixture of at 
least three distinct elements one very tall, long-headed with broad 
deeply depressed nose, dominant in the north; another also tall 

1 The Anthropology of the* North American Indian^ in Memoirs of the 
International Congress of Anthropology > Chicago, p. 38. 


and long-headed with very broad but less depressed nose, domi- 
nant in the south, and a third round-headed, very short and hairy, 
whose domain lies about the equator l . 

In Asia analogous cases occur in Afghanistan, where the 
vigorous Hazaras and Aymaks are of Mongolo- 
Persian descent; in Kashmir, where the Baltis, 
who give their name to the province of Baltistan, 
are described by Major Biddulph and others as an excellent fusion 
of Mongols and Aryans 8 ; in India generally, where masses of 
Dravidian and Kolarian aborigines have benefited immensely by 
their union with the Aryan intruders some thousands of years ago ; 
in Cochin-china, where the Franco-Annamese half-breeds known 
as Minh-huongs are steadily increasing in numbers and displaying 
qualities of a sterling character 8 . Most of the Philippine Islanders 
are the outcome of diverse interminglings, in which the Malay, 
Negrito, Chinese, Japanese, Spanish and perhaps Polynesian 
elements are variously represented. But there is no lack of 
vitality in these mestizos, who have certainly increased in 
numbers under the Spanish rule. 

In Malaysia many of the so-called " Alfuros " are the result of 
crossings between the Malays and Papuans, and 
analogous crossings between Papuans and Poly- 
nesians make up a large part of the population in 
Fiji and Melanesia. A striking instance of the permanently 
fertile union of two extreme types is afforded by the present 
inhabitants of Pitcairn Island in the South Pacific. In 1789 the 
mutineers from the Bounty o English sailors, 

TIL*. < D;*.__i_ r% -^ ' 

island m 6 male and 15 female Tahitians settled on this 

island. But through constant strife and bloodshed 

these were reduced in 1793 to four Englishmen and ten Tahitian 

women. Since then, peace having been restored, the community 

1 U Anthropologies MayJune, 1890, p. 294. 

2 " Les Baltis possdent tout a la fois la patience et la t^nacite du Mongol et 
Tintelligence &eve"e avec Tesprit d'initiative qui caracterisent 1'Arya" (J. van 
den Gheyn, Rev. des Questions Scientifiques, July 1883, p. 5). 

3 Of these Minh-huongs M. Morice tells us that they "deviennent de plus 
en plus nombreux, resistent bien le climat...les enfants, fort gentils, ont le nez 
un peu camus, les cheveux detains, et le teint un peu plus clair que les 
indigenes" (Bui. d. I. Soc. d'Anthrop., Feb. 1875). 


began to flourish, increasing in 1825 to 66, and in 1891 to 120. 
These islanders are a robust, active race, of dark complexion but 
pleasing expression, and very intelligent. Their steady expansion, 
as shown by the colony founded by them on Norfolk Island, 
establishes their permanent eugenesis. In general it may be said 
that in the South Sea Islands it is the full-blood Polynesian 
natives that are disappearing (Maori, Hawaii, Samoans, &c.), 
while their place is being taken by half-breeds of all kinds. All 
these facts, which might be multiplied indefinitely, fully justify Dr 
Robert Dunn's statement that "half-castes very generally combine 
the best attributes of the two races from whence they originate 1 ." 

It may be concluded on inductive evidence that all the 
Hominidae are, and always have been, permanently 
fertile with each' other. Eugenesis is the norma, { 1 j h ^ 1 p ^ io " 
and to it must in fact be attributed the present conclusive 
endless varieties of mankind, which may be said to potygVnists. 
have almost everywhere supplanted the few original 
fundamental stocks. The argument in favour of the specific unity 
of these stocks may be summed up with Prof. E. B. Tylor, who 
remarks that "the opinion of modern zoologists, whose study of the 
species and breeds of animals makes them the best judges, is 
against the view of several origins of mankind, for two principal 
reasons. First that all tribes of men, from the blackest to the 
whitest, the most savage to the most cultured, have such general 
likeness in the structure of their bodies and the working of their 
minds as is easiest and best accounted for by their being descended 
from a common ancestry, however distant. Second, that all the 
human races, notwithstanding their form and colour, appear 
capable of freely intermarrying and forming crossed races of every 
combination, such as the millions of mulattos and mestizos sprung 
in the New World from the mixture of Europeans, Africans and 
native Americans ; this again points to a common ancestry of all 
the races of men. We may accept the theory of the unity of 
mankind as best agreeing with ordinary experience and scientific 
research 9 ." 

1 Unity of the Human Sfecies (physiological and psychological evidence), 
1861, p. 5. 

3 Anthropology, p. 5. 


From the now universally accepted doctrine of correlation of 
parts, Prof. Kollmann draws another argument in 
tomfcaUeVt. support of the unigenist doctrine against poly- 
genist views. After referring to Cuvier's statement 
that from a single bone it is possible to determine the very species 
to which an animal belongs, because every bone stands in such a 
relation to every other, that from its characteristics the characters 
of all the others may be inferred, he adds : " Precisely on this 
ground I have mainly concluded that the existence of several 
human species cannot be recognised; for we are unacquainted 
with a single tribe, from a single bone of which we might with 
certainty determine to what species it belonged 1 ." 

Driven from the physiological and anatomical grounds, poly- 

genists have taken refuge in a philological argument, 

genlst HivT" which they consider unanswerable, possibly because 

guistic anthropologists have hitherto considered it not 

argument. . ._ - . . . . 

worth answering. Of course special anthropology, 
which deals with man only as a member of the zoological series, is 
not called upon to discuss linguistic questions at all. But they 
cannot be overlooked by the ethnologist, who has to study man 
and all his faculties, of which articulate speech is the most 

Abel Hovelacque concludes his Science of Language with the 

remark that "the ascertained impossibility of re- 
independent . ..... ... ...... 

stock Races ducing a multiplicity of linguistic families to a 

in f dlpendeilit common centre is for us sufficient proof of the 
stock Lan- original plurality of the races that have been 


developed with them 20 ; and elsewhere: "If the 
faculty of articulate speech constitutes the sole fundamental 
characteristic of man, and if the different linguistic groups known 
to us are irreducible, they must have taken birth independently 
and in quite distinct regions. It follows that the precursors of 
man jnust have acquired the faculty of speech in different localities 
independently, and have thus given birth to several races of man- 
kind originally distinct. ... Had man acquired this faculty in one 

1 Ueber pithekoide Formen in dem Gesichtssckadel, in Correspondenz-Blatt of 
the German Anthrop. Soc., Nov. 1883, p. 164. 

2 English ed. 1877, P- 3 [1 


way only, language would have remained substantially the same to 
the present time, or at least we should detect in all languages 
some traces of their common descent " (ib. p. 304). So also Fr. 
M tiller in his AUgemeine Ethnographic, and other polygenists, who 
confidently argue from fundamentally distinct stock languages to 
fundamentally distinct stock races evolved in different geographical 

But the inference is based on a tremendous fallacy, which per- 
vades an immense number of ethnological treatises, 
and which does not appear to be anywhere ade- 
quately dealt with. The irreducible stock languages 
are unquestioned, and Mr J. W. Powell enumerates as many 
as fifty-eight for the United States and Canada alone 1 . In the 
rest of the Continent there must be at least as many 

. , . Specific 

more, or, say, at an extremely moderate estimate, unity un- 

one hundred for the whole of America. Are we 
therefore to conclude that there are also at least a stock Lan- 
hundred stock races, a hundred distinct species of guages ' 
the Hominidse in the New World where nevertheless such remark- 
able physical uniformity prevails? And if so, how were they 
evolved in a region, where there are not even any anthropoid apes 
higher than the Cebidse, from which no sane zoologist would 
attempt to trace the ascent of man ? In Australia there are not 
even any Cebidae ; no apes or monkeys of any kind, no half-apes 
or lemurs, no placental mammals, except a few species of bats and 
rodents, but there is at least one stock language. Is therefore the 
race that speaks it to be derived from bats or rodents, or perhaps 
marsupials ? For the geological record shows that in this region 
there never have been any higher mammals except the dingo of 
recent introduction. Here therefore the polygenists must give up 
the problem, or else fall back on direct creation, or perhaps on 
Agassiz' exploded hypothesis of several distinct pairs of " proto- 
plasts," with radiation of species from several distinct centres. 
And all this to avoid the comparatively easy transitions from one 
variety to another of the Hominidse. 

In other parts of the Eastern Hemisphere, such as Sudan, 

1 Indian Linguistic Families, &c., Washington, 1891. 


Caucasia, Malaysia, stock languages are reckoned by the score, 
[n some districts of Caucasia, the "Mountain of Languages," as 
t has been called by the Arabs and Persians, almost every upland 
ralley has its distinct form of speech, and although a few of these 
have been traced to a common source, many have hitherto resisted 
ill the attempts of philologists to classify them in family groups. 
But the inhabitants of these valleys all belong physically to the 
same great Caucasic division of mankind, of which they are in fact 
typical members. Similar relations prevail in the Minahasa 
district at the extremity of the northern peninsula, Celebes, 
where in a small tract some 60 miles by 20 over a dozen 
different languages are spoken. "Some of these may perhaps be 
more or less dialectic, but the majority are said to be quite 
distinct, and the people of the different tribes cannot make 
themselves understood except through the medium of Malay, 
although, perhaps, their villages may be within three miles of one 
another 1 ." In these regions the absurdity of the argument that 
infers stock races from stock languages is thus seen in its full 
force, and the truth of the somewhat trite saying that quod nimis 
probat nihil probat (" what proves too much proves nothing ") is 
strikingly illustrated. It follows, as will more fully appear farther 
on, that there is no necessary relation at all between race and 
speech. In other words, however useful as a factor in determining, 
or helping to determine, the affinities of various races one to the 
other, language has no bearing whatever on the question of the 
original unity or diversity of mankind. 

Nevertheless the absolute irreducibility of the stock languages 
u . ,. is a difficulty, to account for which Prof. Sayce 

which are to J ' J 

be otherwise amongst others suggests that " man was speechless 
explained. w hen the leading races were differentiated from each 
other." But to this the same fatal objection still applies, for on 
this assumption there would be needed not one homo alalus 
primigenius, as postulated by Haeckel, but as many speechless 
precursors as there are and have been stock languages in all parts 
of the world. Probably ten times as many stock languages have 
perished during the long ages since the evolution of speech as still 

1 Dr F. H. H. Guillemard, Australasia (Stanford series, new issue), p. 291. 


survive. Hence it would be logically necessary -to assume that 
this marvellous evolution took place not ten or twenty but many 
hundred times in various regions of the globe. A much less 
violent assumption is the common sense view, not that every 
distinct language represents a distinct race, but that the several 
distinct races must have evolved within themselves a greater or 
less number of distinct languages, that is, of languages which have 
diverged from a common source so far as to become true species 
or even genera and orders, while the races themselves have 
remained mere varieties of a single species. 

No true evolutionist can have any difficulty in accepting this 
view. Linguistic are far more variable than animal or vegetable 
forms, and in anthropology it is a generally accepted principle 
that speech changes more readily and more rapidly than physical 
types. Hence it is more easy to conceive all the present linguistic 
orders deriving from an original germ or inorganic state of primi- 
tive speech, than all the present animal and vegetable orders 
deriving from original animal and vegetable germs. The only 
difference is that the biological series are proved by palaeontology, 
whereas the early linguistic series must necessarily be postulated, 
because extinct forms of speech leave no fossils behind them. 
Historic languages, however, leave documents, and some of these 
documents, such as the Hindu Vedas, reveal an enormous 
divergence in the course of a few thousand years within the limits 
of a single linguistic family. Compare, for instance, modern 
English with Sanskrit, Zend or Homeric Greek, all members of 
the Aryan group. From what has taken place in this relatively 
short historic period, any extent of divergence may be conceived 
as possible, and indeed necessary, during the immeasurably 
longer prehistoric period, until a stage is reached when no 
resemblance at all will be perceptible between the primitive and 
later Aryan tongues. They will have become radically distinct, 
that is, stock languages. 

Thus the existence of the present stock languages is no argu- 
ment at all for the disparity of the human family j while on the 
other hand the fact that every single member of that family is a 
speaking animal supplies perhaps the very strongest argument for 
the specific unity of all its branches. Waitz aptly remarks that 


"inasmuch as the possession of a language of regular gram- 
matical structure forms a fixed barrier between 

The Mono- 

genist view man and the brute, it establishes at the same time 
and confirmed a near relationship between all peoples in psychical 
bytheuni- respects....In the presence of this common feature 

versal diffusion r . \ 

of articulate of the human mind, all other differences lose their 
speech. importance 1 ." And he quotes Pott as saying that 

"if theology feared that an original difference of language might 
implicate the original unity of the human species (which by no 
means follows), the science of language restores to theology the 
psychical unity of mankind, compared with which the physical 
unity must yield in importance " (tb.) This argument in favour of 

unity, based on psychological grounds, was urged 
Argument. w ^ muc h force and eloquence by Dr Prichard, who 

pointed out that "the same inward and mental 
nature is to be recognised in all the races of men. When we 
compare this fact with the observations, fully established, as to the 
specific instincts and separate psychical endowments of all the 
distinct tribes of sentient beings in the Universe, we are entitled 
to draw confidently the conclusion, that all human races are of 
one species and one family 2 ." 

Blumenbach, true founder of scientific anthropology, has 
The uestion summe< ^ U P ^ e whole question from the physical 
summed up by standpoint in words which have lost nothing of their 
Blumenbach. f orce s { nce t h ev were penned a hundred years ago. 
He asks whether everywhere in time or place mankind has con- 
stituted one and the same, or clearly distinct species ; and he con- 
cludes: "Although between distant peoples the difference may 
seem so great, that one may easily take the inhabitants of the Cape 
of Good Hope, the Greenlanders and Circassians for peoples of so 
many distinct species, nevertheless we shall find, on due reflection, 
that all, as it were, so merge one in the other, the human varieties 
passing gradually from one to another, that we shall scarcely if at 
all be able to determine any limits between them. Hence those 
varieties of mankind have proved extremely arbitrary both in 

1 Anthropology, p. 273. 

a Natural History of Man, p. 488. 


number and description, which have been accepted by distin- 
guished men 1 ." The last remark will receive its full justification 
in the next chapter. 

1 "Sintne fuerintne omnis aevi omnisque gentis homines unius eiusdemque 
diversseve plane specie!... Quamquam tanta inter remotiores gentes interesse 
videatur differentia, ut facile Capitis Bonse Spei accolas, Groenlandos et Cir- 
cassios pro tot diversse speciei hominibus habere possis, re tamen rite pensitata, 
ita omnes inter se confluere quasi, et sensim unam in alteram transire hominum 
varietatem videbis, ut vix ac ne vix quidem limites inter eas constituere poteris. 
Maxime arbitrarice ideo et numero et definitione evaserunt quas cl. viri 
receperunt generis humani varietates " (De generis humani varietate naliva, 
I795 P- 40)- 

K. II 



Difficulties of defining, and determining the number of, the primary human 
varieties Schemes of the first systematists : Bernier; Linne; Blumenbach; 
Cuvier; Virey; Desmoulins ; Bory de Saint- Vincent ; Morton; Gliddon 
and Agassiz; Latham; Carus; Peschel The Philologists The Ethno- 
logists : Buffon ; Prichard The Anatomists : Geoffroy Saint- Hilaire ; 
Retzius ; Broca ; Virchow ; Mantegazza ; Barnard Davis ; Rolleston ; 
Flower; Cope Recent Schemes : Haeckel's; de Quatrefaq;es's; Huxley's; 
Broca's; Fr. Miiller's ; Deniker's; Flower and Lydekker's General 
remarks on these GroupingsElements of Classification : Physical and 
Mental Characters Physical tests of Race : Colour of the Skin Colour 
and Texture of the Hair The Beard ; Hirsuteness Shape of the Skull 
Cephalic Indices Tables of Dolicho-, Mesati- and Brachycephali 
Gnathism Facial Index Table of Sub-nasal Prognathism The Denti- 
tion The Nose: Nasal Index Colour and Shape of the Eye The 
General Expression Stature : Tables of Heights Other Physical Factors. 

FROM the foregoing considerations it appears that the Homi- 
nidae constitute a family group, that is, a group connected, how- 
ever distantly, by the ties of blood derived from a common 
pliocene precursor. It further appears that several distinct 
members of the group were already established in pleistocene 
times in every part of the then habitable globe. It follows that 
the present races of mankind are to a certain extent of diverse 
origin, that is to say, descend in diverging, converging or 
parallel lines from their several pleistocene pre- 
cursors, without anywhere developing specific differ- 
ences. But to this very fact of their relatively 
the primary'" close kinship is due the great and admitted difficulty 
"" of determining the number and character of the 

existing primary groups. It was seen that because 
the Equida form so many true species, systematists find it an 


easy task to define and describe their main divisions. On the 
other hand the grouping of the Canidae, which form varieties only, 
presents almost insuperable difficulties to classifiers. 

In this respect the position of the Hominidae is entirely 
analogous to that of the Canidae. All being fertile inter se, 
although possibly in different degrees, and several having early 
acquired migratory habits, endless new varieties have constantly 
been formed since remote prehistoric times, both by segmen- 
tation of early groups, and by countless fresh combinations of 
already established varieties. Outward modifying influences must 
have been brought into play as soon as the first-named groups 
began to migrate from their original homes, and such influences, 
intensified by the climatic changes accompanying the advance 
and retreat of glacial phenomena, would increase in activity 
according as the primitive tribes spread farther afield. To these 
influences of the surroundings were soon added the far more 
potent effects of interminglings seen to be at work already in 
neolithic times, and thus the development of fresh sub-varieties 
of all sorts proceeded at an accelerated rate. This process 
has necessarily continued down to the present time, resulting in 
ever-increasing confusion of fundamental elements, and blurring 
of primaeval types. Hence it is not surprising that many ethno- 
logists should accept as a truism the statement that "there are 
no longer any pure races in the world 1 ." 

To this ethnical confusion, which has been traced back to the 
megalith-builders, and even to the Furfooz, Finale, 

& , , ., . ' Schemes of 

and other cave-dwellers, must be attributed the the first 
amazing diversity of opinion that has prevailed and s y stematlsts - 
still prevails amongst anthropologists, even as regards the number 
of the primary divisions of mankind. The first 
serious attempt at a systematic grouping of the 
Hominidae has been accredited to F. Bernier (1625 88), who 
distinguished (1672) four radical types: the European white, the 

1 "Le scul substratum sur lequel nous pouvons operer, les divers peuples, 
nations, peuplades t tribus &c., tels qu'ils sont actuellement repartis sur la terre, 
ne sont que les melanges d'elements souvent tres heterogenes. La phrase : ' II 
n'y a plus de races pures sur la terre,' est devenue un cliche " (J. Deniker, 
Bui. d* 1. Soc. d'Anthrop., June 6, 1889, p. 322). 

II 2 


African black, the Asiatic yellow, and the northern Lapp ! Then 
came the great systematist Linne' (1738 83) with 
his Homo mofislruosus. Homo ferns, and Homo 
sapiens. The Homo ferus, being dumb and covered with hair, 
answers somewhat to Haeckel's Homo alalus, while the group 
Homo sapiens comprises four varieties : the fair-haired, blue-eyed 
and light-skinned European; the yellowish, brown-eyed, black- 
haired Asiatic ; the black-haired, beardless, tawny American ; the 
black, woolly-haired, flat-nosed African. Biumenbach (1752 
1840) followed (1775) with his five varieties bearing 

Biumenbach. ' , / . , , ^ - 

a nomenclature that still largely persists : Caucasic, 
Mongolic, Ethiopic, American and Malay. But Biumenbach later 
(1795) fell back on Linne's four varieties which, however, he 
distributed somewhat differently, assigning to the Caucasic most 
of Europe, Cis-gangetic Asia and the region stretching north- 
wards from the Amur basin ; to the Mongolic Trans-gangetic Asia 
north to the Amur "with the islanders and great part of the 
Austral lands"; to the Ethiopic Africa; and to the American all 
the New World except the northern coastlands, that is, the 
Eskimo domain, which he includes in the Mongolic division 1 ." 
Then ensued a period of orthodox reaction against the 

Lamarckian ideas headed by Cuvier (1773 1838), 

who held by fixity of species, but inconsistently 
admitted three races, the Caucasic, Mongolic and African, sup- 
posed to answer to the biblical Japhetic, Semitic and Hamitic 
families. This of course caused a great outcry, and in fact was 
the starting-point of the monogenist and polygenist theories, 

which were discussed in the last chapter. In 1801 

Virey (1775 1840) reduced Cuvier's three divi- 
sions to two distinct species, white and black, each with three 
main races or sub-species, which again comprised a number of 
secondary groups. But this could not satisfy thorough-going 

polygenists, such as Desmoulins, who started eleven 
Desmouhns. h uman S p ec i es i n 1825, and next year raised them 
nt, to s i xteen > Bory de Saint- Vincent, who in 1827 

discovered fifteen species, including such nebulous 

1 Op. cit. p. 42. 


groups as "Scythians," " Neptunians," "Columbians"; lastly the 
American school, which in the hands of Morton, 


Gliddon, Knox, Agassiz and others brought about GHddon, 
an inevitable reaction by threatening to increase B assiz 
the number of species indefinitely. Other groupings, which were 
marked by greater sobriety, and which still possess some historic 
interest, were those of Hamilton Smith (Caucasic, Mongolic, 
Tropical); Latham (Japhetic, Mongoloid, Atlan- L atham 
tides); Karl G. Carus (four divisions somewhat 
phantastically named Nachtmenschen, "Night-men," the Negro; 
Tagmenschen, "Day-men/* the Caucasian; ostliche 
Ddmmerungsmenschen^ "Men of the eastern twi- 
light," Mongolo-Malayo-Hindu peoples; and westliche Dammer- 
ungsmenscheri) " Men of the western twilight," the 
American aborigines); lastly Peschel (Australian 
with Tasmanian, Papuan, Mongoloid with Malayo-Polynesian and 
American, Dravidian, Hottentot with Bushman, Negro, Mediter- 
ranean, i.e. Blumenbach's Caucasian). 

A fresh element of confusion, which still clings to ethno- 
logical studies, arose out of Frederick Schlegel's 
little treatise on the "Language and Wisdom of the 
Hindus" (1808), which was later declared by Max 
Mu'ller to have revealed a new world, and to have shown what 
unexpected services Anthropology might derive from the science 
of language. Unfortunately these services were pushed too far 
when philologists entered the field, and claimed to hold in 
language the key to the solution of all ethnological problems. 
This again led to another reaction, caused especially by the 
attempt to identify race and speech, and to set up as many 
independent physical as there are independent linguistic groups, 
as discussed in the last chapter. When it was seen that such 
views led, like those of Nott, Gliddon and Knox, to an un- 
limited number of human species and varieties, a violent divorce 
took place between philology and ethnology, a divorce which will 
be dealt with farther on with a view to a possible reconciliation 
of the two schools. 

Meanwhile the way had been prepared for a more rational 
treatment of racial diversity by Dr James Cowles Prichard, who 


not without reason is by many regarded as the true founder of 
ethnology as a distinct branch of general anthro- 
pology. At least he may share this honour with 

Buffon, BufTon, who so early as 1749 had undertaken 

rTicnara. * 

I* Histoire Complete de rHomme, as a part of his great 

work on the Animal Kingdom (1749 88). Both of these great 
writers avoided, perhaps wisely, any systematic groupings, but 
brought to bear a great store of learning, combined with much 
acute reasoning, on the natural history of the various divisions of 
mankind, as they presented themselves in their several geographic 
areas. But while Buffon was mainly descriptive (Ethnography), 
the comparative method is conspicuous in Prichard (1785 1848), 
whose writings (Eastern Origin of the Celtic Language; Physical 
History of Mankind, &c.) are consequently of a strikingly ethno- 
logical character, and possess great permanent value. 

His Crania of the Laplanders and ^inlanders, continued by 

Theanato- tne more solid work of the elder Retzius in the 

mists: same field, gave a fresh impulse to craniologicai 

Geoffrey Saint- ,. ,.,,,, , , , . , f 

Hiiaire, studies which had already been cultivated by 

Retzius. Morton, and on which Geoffrey Saint-Hilaire based 

his four fundamental types : orthognathous, eurygnathous, progna- 

thous and eury-prognathous (1858). Thus were laid the foun- 

dations of the comparative study of the Hominidae based on their 

physical characters, a line of inquiry which in the hands of 

Broca, de Quatrefages and Hamy (Crania Ethnica), 

Broca, Topinard, Virchow, Kollmann, Mantegazza, Pruner 

v?rchowf CS ' Bey, Barnard Davis, Beddoe, Huxley, Thurnam, 

> Turner, Rolleston, Flower, Macalister, Garson, 

Davis, Cope and others, has led to fruitful results. On 

Rolleston, 11-11 /- i 

Flower, Cope, these physical characters, for the most part irre- 

spective of speech or other mental qualities, were 

established fresh groupings, which have entirely superseded the 

more extravagant polygenist classifications, while showing a 

general tendency to revert to Linnets and Blumenbach's primary 

divisions in various more or less modified forms. 

scheme?. As reference is constantly made in ethnological 

Quatiefages's. writin g s to one or more OI " tnese groupings, a brief 

summary is here appended of the more important. 



Ulotriches jLophocomi (Tufted) : Papuans ; Hottentots. 
(Woolly-haired) lEriocomi (Fleecy) : Kafirs; Negroes. 

Euthycomi (Straight) : Malay, Mongol, 


American, Arctic, 

Euplocomi (Curly) : Dravidas, Nubians, 



, (White or Caucasic \ ^ , ^ , 
Souche L T _, .. Boughs, Branches, 

/i> *\ i Yellow or Mongolic^ _, % ' ' 

(Root) 1 _ . . 6 . tamihes, Groups, &c. 

v ' (Negro or Ethiopic J 


(LEIOTRICHI, "smooth-haired," and ULOTRICHI, "woolly- 
haired," adopted from Bory de Saint-Vincent) 

Ulotrichi : yellow-brown to jet-black ; hair and eyes dark ; 
mostly long-headed ; Negro, Papuan. 

Leiotrichi : (a) Australoid, dark skin, hair, and eyes ; hair long 
and straight ; prognathous ; Australians, the blacks of the Dekkan. 

(b) Mongoloid, yellow-brown, or reddish-brown ; dark eyes ; 
long, black, straight hair; mesaticephalous ; Mongols, Chinese, 
Polynesians, Eskimo, Americans. 

(c) Xanthochroid, fair skin, blue eyes, abundant fair hair ; 
mesaticephalous ; Slavs, Teutons, fair Kelts. 

(d) Melanochroid, pale skin, dark eye, long black hair ; 
Iberians, Berbers, dark Kelts. 

1 Classification des Races Humaines^ p. 798. Here the terminology is 
defective, the word group being indefinite ^ while all the others are definite. An 
attempt is made in working out the scheme to correlate the three fundamental 
linguistic to the three fundamental physical groups. But the result is vitiated 
by the prevailing misconception regarding the so-called " Monosyllabic Lan- 



I. STRAIGHT-HAIRED: i. dolicho, Eskimo; 2. brachy, (a) red, 
Prairie Indians ; (b) olivaster, Mexican, Peruvian ; (c) yellow, 
Guarani, Samoyede, Mongol, Malay. 

IT. WAVY or CURLY-HAIRED: i. dolicho, (a) blonde, Cim- 
merian, Scandinavian, Anglo-Saxon; (b) brown, Mediterranean 
(Basque, Corsican, Berber) ; Semite ; (c) black, Australian, Indo- 
Abyssinian ; (d) red, Fulah, Red Barabra (Nubian) ; 2. brachy, 
blonde, Finn ; chestnut, Kelt, Slav ; brown, Iranian, Galcha. 

III. WOOLLY-HAIRED : i. dolicho, (a) yellowish, Bushman, 
(b) black, Oceanic, Papuan; Africa, Kafir; 2. brachy, Negrito. 


I. WOOLLY : i. Tufted, Hottentot, Papuan ; 2. Fleecy, Negro, 

II. SMOOTH : i. Straight, Mongol, Arctic, American, Malay, 
Australian ; 2. Wavy, curly, Dravidian, Nubian, Mediterranean, 


This remarkable scheme 1 needs a word of explanation. On 
the assumption that every ethnical group results from a fusion 
of two, three or more "races," the characters of each of which 
are persistent, it follows that every such group must contain 
within itself two, three or more distinguishable strains, here called 
"types." Thus there are more" types" than " races," which at 
first sight sounds paradoxical, and from thirteen racial groups 
are in fact evolved " thirty types," set forth in a scheme primarily 
based on the different textures of the hair. The races themselves 
are further disposed in a space of two dimensions (three being 
impracticable) in order the better to show their mutual affinities, 
which could not be done in the usual linear arrangement 8 . 
Thus ; 

1 Bui. d. I. Soc. d'Anthrop. June 1889. 

2 ** Pour bien presenter ces affinites, il faudrait disposer les groupes suivant 
les trois dimensions de 1'espace, ou du moins sur une surface ou 1'on a la 
ressource de deux dimensions. C'est ce que j'ai essay e de faire sur le tableau 
suivant ou les races sont disposees approximativement d'apres leurs affinites 






I. Ethiopian, Negroid or Mdamsian : Negro, Negrillo, Bush- 
man, Papuan (Oceanic Negro), Australian. 

II. Mongol, Xanthous or Yellow : Mongols, Malays, Poly- 
nesians, Americans somewhat doubtfully as an " aberrant" 

III. Caucasic or Eurafrican : Huxley's Xanthochroi and 

None of these schemes profess to be more than tentative 
efforts at a satisfactory classification, where the initial General Re 
difficulty lies in the fact that the groups themselves marks on these 
are already mixed. Some are based on positively 

rou P m & s - 

dites naturelles" (ib. p. 328). No doubt the unfortunate use of the word 
" type " has damaged this scheme, which however in the details gets entangled 
in several incongruities, due to the difficulty of separating fused or juxtaposed 
strains from the different racial groups. 

1 Introduction to the Study of Mammals, p. 743. 


erroneous data, such as Huxley's, which treats the Eskimo as 
mesaticephalous (intermediate between the long and the round 
heads), whereas they are, next to the Fijian Kai-Colos, the most 
dolichocephalic of any race. So with Haeckel's and Fr. M tiller's 
Hottentots and Papuans, grouped together on the ground of the 
" tufted " hair growing in separate bunches, which, however, is 
characteristic, not of the Papuans, but of the Hottentots and 
Bushmen alone. Even in these the hair grows uniformly over 
the scalp, as amongst all other races, and not in isolated tufts 
with intervening bald spaces, as is often asserted. Owing to the 
peculiar clumpy growth such spaces merely simulate baldness, 
though in old age they actually do become bald, thus giving rise 
to the prevailing mistake 1 . Broca brackets Fulahs and Nubians 
together because of their common "red" colour, neither being 
red, and both being in other respects quite distinct the Nubians 
being originally Negroes from Kordofan mixed with Bejas and 
other eastern Hamites, and the Fulahs originally and still partly 
Saharan Hamites, mixed here and there with Mandingans and 
other West Sudanese Negroes. Fr. Miiller appears to be primarily 
responsible for this " Nuba-Fulah Family," constituted on a 
linguistic base, the two languages being fundamentally distinct 2 . 
Again, what is to be made of the expression " Indo-Abyssinian," 
or even " Abyssinian " at all as an ethnical term ? The very word 
(Habeshi) means " mixed," and in African ethnology " Abys- 
sinian " conveys no more meaning than does " Hungarian " in 
European ethnology; both are national not racial designations, 
and as a Hungarian may be a Magyar, a Slav, a Rumanian or a 
Teuton, so an Abyssinian may be a Hainite (Agao and others), 
or a Semite (Tigre and others). 

1 " Ce dernier caractere [cheveux laineux] atteint son maximum dans les 
chevelures dites en grains de poivre ['peppercorn' growth] que Ton a cru 
longtemps pousser par touffes isolees. De nouvelles recherches et une obser- 
vation tres precise de M. Topinard ont montre qu'il n'en est rien " (De Quatre- 
fages, op. cit. p. 203). See also J. Deniker, Rev. d'Anthrop. 1883, P- 49<> 
where the error is explained by the fact that "tres sou vent les cheveux des 
Fapous [Hottentots] s'enchevetrent et forment de petites boules simulant 
des Unifies separees." 

2 A. H. Keane, Ethnology of Egyptian Sudan, p. 16- 


From a general survey of the various schemes, it appears that 
special, if not paramount, importance is given by 
these systematists to the three elements of com- 

plexion, character of the hair, and shape of the skull, physical and 

... i i * r T i mental charac- 

And, in general, physical features are relied on, not t ers. 
merely in preference to, but to the total exclusion 
of mental qualities. Yet in determining the relative position of 
ethnical groups these cannot be overlooked, else ethnology re- 
mains merely a sub-branch of special anthropology, which as seen 
confines itself to the human anatomy, and disregards the intel- 
lectual side, in virtue of which alone the Hominidae constitute an 
entirely separate division of the animal series. Nor can it be said 
that the mental endowments are all alike, and consequently 
useless for schematic purposes. On the contrary they show far 
greater diversity than do the physical qualities, as is evident from 
the single fact that, as seen, languages form distinct species and 
genera, while the various human groups constitute varieties only. 
Hence due account will here be taken of the mental as well as of 
the physical characters, as criteria of racial affinities. 

Precedence may be claimed for colour, at least as the element 
which occurs first to the observer, and on which, 
probably for that reason, the first groupings were 
determined. Nevertheless we are warned by Linne' 
himself not to trust too much to this character : ne nimis credecolori, 
. and physiology now tells us that it is mainly, if not 
essentially, a question of climate and, quite possibly, 
diet 1 . It appears that the pigment, or colouring 
matter, situated chiefly in the rete mucosum or lower layer of the 
cuticle, which was formerly supposed to be peculiar to the Negro, 
is really common to all races, only more abundant and of darker 
hue in the Negro, the Papuan, Australian and Oceanic Negrito. 
Nor is there any necessary correlation between this darker hue 
and other Negro characters, as appears from its presence in 
many Somal, Galla and other Hamitic and even Semitic groups 
of quite regular features (see p. 382). Waitz (op. cit. pp. 46 
52) adduces many examples to show that "hot and damp 

1 " Principem tamen inter omnes nigredinis causas locum tenebit clima, 
soils aerisque potentia cum vitse genere " (Blumenbach op. cit. p. 50). 


countries favour the darkening of the skin," and that the same 
race tends to be much darker in low, marshy districts than 
on the neighbouring uplands. Lepsius, a good observer, de- 
clares somewhat emphatically that the hotter the climate, the 
darker is the colour of the Negro ; he adds that, proceeding 
from Africa eastwards the isothermal line of greatest heat inter- 
sects the regions in southern Asia, which are inhabited by the 
darkest peoples of that Continent 1 . There may be some exaggera- 
tion in this statement, and there are certainly many apparent 
exceptions to the general law regarding the direct relation of 
heat to colour. But the exceptions are probably either due to 
local causes, or to the absence of one or other of the factors which 
combine to darken the pigment. Thus Schweinfurth (Heart of 
Africa) attributes the reddish hue of the Bongos and other 
Negroes of the hot, moist White Nile basin to the ferruginous 
nature of the laterite soil, and the same cause appears to have 
produced the same result amongst the A-Zandeh (Niam-Niam) of 
the Welle valley. 

In America all shades within certain limits seem to be inter- 
mingled irrespective either of latitude, temperature, or relief of the 
land. Thus in Bolivia are found in juxtaposition the coppery 
Maropas, the dark brown Aymaras, the yellowish Moxos and the 
light Mosetenos, Siriones and Guarayos 8 . So in Australasia the 
yellow-brown Malays living about the equator present a striking 
contrast to the almost sooty black Tasmanians of the south 
temperate zone. But physical as well as moral characters are the 
outcome not of one or two but of many causes acting simul- 
taneously on the organism, which cannot escape either from its 
environment or from its own tendencies. Hence such seeming 
discrepancies are to be attributed either to descent (dark peoples 
migrating to cold, light to warm regions), or to various local 
circumstances and other influences, such as dryness, moisture, 
food, aspect, altitude or flora (herbaceous or arboreal) of the land, 
by all of which the complexion may be diversely affected, and 

1 Nubische Gram. Einleitung. This isothermal does not coincide in Africa 
with the equator, but is deflected in the north to about 1-2 15 N., precisely 
where are found the Wolofs and other Negroes of the deepest dye. 

a A. H. Keane, Indians (American) in Encyc. Brit., Qth ed. 


mere temperature largely neutralised. The Negro migrating from 
the moist tropical zone northwards and southwards, or trans- 
planted to the cooler regions of America, will for an indefinite 
period retain his inherited dark colour, which, whatever its origin, 
has to a large extent become a racial character. On the other 
hand the Semitic and Hamitic inhabitants of the intensely hot 
but also intensely dry regions of Arabia and the Sahara, are of 
distinctly light complexion, not perceptibly darker than many 
South Europeans. 

It is important to note that the palms and soles of the Negro 
are never black, but always yellowish, that the dark pigment is 
wanting in the Negro foetus, and that Negro children are born "of 
a light grey colour" (Waitz, p. 99). Hence it might be inferred 
that the dark colour, with which a thicker skin is correlated, is a 
later development, an adaptation of the organism to a hot, moist 
malarious climate, in which the Negro thrives and the white man 
perishes. Thus colour taken alone cannot be regarded as an 
entirely trustworthy test of race, the less so that even blackness l is 
not an exclusively Negro character, but common also to many 
eastern Hamites (Agaos, Bejas, Somals, Gallas), and to numerous 
aborigines of India. Nevertheless it is far too important a factor 
to be overlooked, and taken in combination with other characters 
will lead to satisfactory results. Although the transitions, as in 
other physical traits, are complete, there appear to be about six 
primary colours to which all the human groups may be referred, 
as under : 

Black: African and Oceanic Negroes; Australians; Tas- 
manians; some aborigines of India and America; Eastern 

Yellow : Mongols; Indo-Chinese; Japanese; Tibetans; some 
South- Americans ; Bushmen; Hottentots. 

1 An absolutely black complexion is of extremely rare occurrence in any 
branch of the Negro group. This may be easily seen by comparing the colour 
of the face of the average African or Papuan with that of his hair, which 
is usually intensely black. The skin will always show a lighter, as well as a 
different shade, so much so that a Negro with face and hair of exactly the same 
sombre hue would look like some monstrous lusus natures, or some stage figure, 
such as the Othello whose weak points were detected by Blumenbach during 
his visit to London (1816). 


Brown ; Polynesians ; Hindus ; Plateau Indians of America ; 
many Negritoes ; Fulahs. 

Coppery red: Prairie Indians ("Redskins"). 

Florid white: Northern Europeans; Lapps; Finns; Xantho- 
chroid Caucasians generally. 

Pale white: Southern Europeans; Iranians; many Semites 
and Western Hamites ; Melanochroid Caucasians generally. 

A glance at the foregoing schemes of classification will suffice 
to show that the hair, if not regarded as of more 

Colour and . . " 

texture of the importance than the complexion, has at all events 
hair * steadily risen in favour with systematists. Its for- 

tune, so to say, was made by the classical memoir, " On the human 
hair as a race character, examined by the aid of the microscope," 
read by Dr Pruner-Bey before the Paris Anthropological Society, 
March 19, 1863*. Since then this element, previously little 
attended to, has been made the base or leading character in the 
groupings of some of the most eminent recent ethnologists. The 
reason is that both colour and texture of the hair are found to be 
extremely constant characters, resisting time and climate with won- 
derful tenacity, and presenting remarkable uniformity throughout 
large sections of the human family. Thus all the American 
aborigines from Fuegia to Alaska, as well as most of the Mongo- 
loid, Malay, and Eastern Polynesian peoples, are invariably 
distinguished by the same black, lank, somewhat coarse and 
lustreless hair, round or nearly round in transverse section. No 
other single physical trait can be mentioned which is to the same 
extent characteristic of several hundred millions of human beings 
distributed over every climatic zone from the Arctic to the Ant- 
arctic waters, and ranging from sea-level (Fuegia, Mackenzie 
estuary) to altitudes of 12,000 and even 16,000 feet (Bolivian and 
Tibetan plateaux). So also short black woolly, or at least crisp, 
or frizzly hair, elliptical and even somewhat flat in transverse 
section, is a constant feature of the Negroes, Hottentots, Bushmen, 
Negritoes, Papuans, Melanesians, Tasmanians, in fact of all the 
distinctly dark Negroid populations, say, of 150 million members 

1 An English translation appeared in the Anthrop. Rev. February, 1864. 
The genera conclusions arrived at by this eminent anthropologist have been 
confirmed and extended by the later researches of Topinard and others. 


of the human family. The only important exception are some 
African pigmies, the Wochua amongst others, whose hair is de- 
scribed by Junker as "of a dark, rusty-brown hue." This observer 
adds that " this is certainly one of the most marked peculiarities of 
the race, for the hair of all other Negro peoples, however light- 
coloured they may otherwise be, is always the deepest black 1 ." 
Lastly hair of intermediate types, black, brown, flaxen, red, smoc th 
wavy or curly, and generally oval in transverse section, prevails 
amongst both sections of the Caucasic division, which may now 
be estimated at 700 or 800 millions. Hence the quality of the 
hair has naturally come to be regarded as one of the safest, if not 
the very safest test of racial purity, and Pruner-Bey goes so far as 
to suggest that " a single hair presenting the average form charac- 
teristic of the race might serve to define it," adding, however, that 
"without pretending to this degree of certainty, it is indubitable 
that the hair of the individual bears the stamp of his origin " 

(P- 2 3). 

It might be objected that hair can have only a secondary 
importance, because, unlike the cranium, it is limited in point of 
time, no specimens having survived from the palaeo or neolithic 
eras. But the Egyptian mummies (some of the fourth dynasty) 
show that for at least 6000 years this feature remains unchanged. 
Hence it may perhaps be inferred that the primary divisions of 
mankind were always distinguished by the same texture and colour 
of hair as at present. But it is specially noteworthy that, as pointed 
out by Topinard 2 , the white group comes nearest to the higher 
apes in this respect, the black being the farthest removed, and the 
yellow intermediate. The lanugo of the human foetus would seem 
to imply that the pliocene, or at all events the miocene precursor 
was a furred animal, and fur might easily pass in one direction 
into lank, in another into woolly, crisp, or intermediate types 
(cf. the goat and sheep). By the wavy intermediate forms may 
perhaps be bridged over the otherwise impassable gulf between 
the lank- and the woolly-haired Ho'minidse. Unless the present 
human varieties are studied with reference to a generalised pre- 
cursor, as the Solidungula and other mammalian groups are studied, 

1 Travels in Africa, ill. p. 82. 

8 UHomme dans la Nature, 1891, chap. VI. 


it will be difficult for monogenists to hold their ground against the 

From Pruner's microscopic studies it appears that, apart from 
its colour, the structure of the hair is threefold : i. Short, crisp or 
fleecy, usually called "woolly," elliptical or kidney-shaped in 
section, with mean diameters 20 : 12 in hundreds of millimetres; 
no perceptible medullary tube, and often relatively flat especially 
in Papuans ; colour almost invariably jet black ; characteristic of 
all black races except the Australians and aborigines of India. 
2. Long, lank, of the horse-mane type, cylindrical, hence round 
or nearly so in section, with diameters either about 24, or if 
elongated 27: 23; distinct tube filled with medullary substance; 
colour mainly black or blue black ; characteristic of all American 
and Mongoloid peoples. 3. Intermediate, wavy, curly or smooth ; 
oval in section, with long and short diameters 23 : 17 or 20 : 15 ; 
distinct tube, but empty or diaphanous ; all colours from black 
through every shade of brown to flaxen, red and towy ; character- 
istic of most Caucasic peoples, but in the eastern Hamites and 
some others developing long ringletty curls. Besides the three 
typical transverse sections : 

Flat ellipse Circle Oval ellipse 

(Negro). (Mongol). (European). 

considerable diversity is presented by some hair, whose sections 
take square, triangular, kidney-shaped or other forms, as thus : 

In general the flatter the hair the more it curls; the rounder 
the more stiff and lank it becomes, these two extremes being 
respectively represented by the Papuan (diameters 29 : 10 or 25 : 7) 
and the Japanese (section a perfect circle). It would also appear 
that of all forms the woolly is the most persistent, as well shown 
by the Brazilian Cafusos, Negro and native half-breeds, who are 


mop-headed like many Papuans. A triple hybrid also, figured by 
de Quatrefages (p. 48) and described as "half Negro, a quarter 
Cherokee and a quarter English," has short, crisp, furry-looking 
hair, and it would seem as if in this respect the Negro hair had 
least deviated from the suggested original fur type. 

With the hair of the head is correlated that of the face, of 
which it will suffice here to remark that the beard 
is properly characteristic only of the Caucasic group. 
All American, Mongoloid and Negroid peoples are normally 
beardless, the chief exceptions being the Australians and some 
Melanesians. Fully developed beards, combined with a general 
hirsuteness, occur also sporadically amongst certain isolated groups, 
such as the Todas of Southern India, the Veddahs of Ceylon, and 
especially the Ainus of Japan 1 . The significance of these facts 
will be dealt with farther on. 

With the shape and size of the skull as racial tests we seem to 
enter debateable ground. On its size obviously 
depend the volume and weight of the brain, on sk yn apeofthe 
which, as seen (p. 44), largely but not exclusively 
depends the mental capacity. Hence this factor will best be 
considered in the next chapter dealing with the intellectual 
qualities. With regard to the shape, to which our remarks will 
consequently here be confined, it may be admitted that no physical 
character has been more extensively studied with, on the whole, 
such indifferent results. Hence the emphatic protests that have 
been uttered by Wallace (p. 43), and some other eminent natural- 
ists against craniology as affording trustworthy data for ethnical 
classifications. Even professional craniologists often express dis- 
appointment at the poor returns for the labour expended. Thus 
Topinard, for whom this line of research forms " the first chapter 
in anthropology," is fain to confess that craniology "in its present 
phase is still a science of analysis and of patience, and not yet 
a science of synthesis 2 ." Miklukho Maclay also, finding the 
heads of New Guinea Papuans varying as much as from 62 (ex- 

1 Many of Junker's Wochua dwarfs "had full beards and hairy breasts," 
though his observations "did not confirm the statement that many of these 
pigmies have very hirsute bodies " ( Travels in Africa, in. p. 82). 

2 Anthropology^ p. 206. 

K. 12 


tremely long) to 86 (round), appears at last to have lost faith in 
craniology as a racial test. He asserts in one place 1 that it cannot 
be regarded as a means of distinguishing between Negritoes and 
Papuans, both displaying an obvious tendency towards brachy- 

But it should be noticed that Maclay appears to have measured 
mostly mixed Papuan specimens, and Sir W. Flower has placed it 
beyond doubt that the typical Negritoes are brachycephalous, the 
typical Papuans extremely dolichocephalous. Some other general- 
izations may also be considered as fairly well established, as, for 
instance, that the African Negroes, Hottentots, and Bushmen are 
normally long-headed, as are also the Arabs (Semites), the Berbers 
(Hamites), the Xanthochroid Europeans and the Eskimo, while 
most of the Mongoloid peoples are round-headed, the Malays 
and American aborigines mixed. A general survey of the ascertained 
facts leads to the inference that of itself the shape of the skull is an 
extremely persistent character, but that it becomes easily modified, 
not perhaps by climate or other outward influences, but certainly 
by intermixture. It follows that remarkable uniformity prevails, 
not only amongst the primitive palaeolithic races (all long-headed, 
p. 149), but also amongst many relatively pure living races, such as 
the Galchas, Savoyards and Auvergnats (all round-headed), and 
the Fijian Kai-Colos (all long-headed), these peoples being pre- 
served from contact with their neighbours by their secluded upland 
or insular homes. Hence also mesaticephalous (intermediate) 
forms may have their value in determining the presence of two or 
more ethnical elements, as in America and Malaysia. 

Craniologists generally assume two fundamental types, the 
dolichocephalous or long horizontally, that is, from back to front, 
and the brachycephalous, or approximately round horizontally 8 . 
The types are determined by the so-called cephalic index num- 
bers, that is, the relation of the antero-posterior 
indies* 1 * diameter (measured from the glabella to the farthest 

point of the occiput) to the transverse diameter 

1 Isvestia, 1879, P- 39 quoted in Nature, Nov. 20, 1879. 

2 Gr. 5o\tx<Ss, long; ppaxfa, short, and Ke^aXi}, head. These terms, which 
play such a large part in anthropological works, were introduced by the elder 
Retzius, true founder of craniology. 


from side to side. The former being taken at 100, the latter will 
range from about 60 to 95, increasing with the greater degree of 
brachycephaly, and vice- versa (see formula, p. 426). Excluding 
artificial deformation, the extremes appear to lie between 61 '9 a Kai- 
Colo of Viti Leva, Fiji, measured by Flower 1 , and 98*21 a Mongolian 
of doubtful provenance described by Huxley. This last approaches 
the perfect circle, which is never presented by the normal head, 
though exceeded (103, 105?) by pathological or deformed speci- 
mens. Most peoples are mesaticephalous, that is to say, they are 
of mixed descent, and it has been seen that the intermingling 
began in neolithic times. Hence it is that, speaking broadly, the 
horizontal index is now applicable less to the primary than to the 
secondary divisions of mankind 2 . The statement, for instance, 
that the African Negroes are normally dolichocephalic, is subject 
to numerous exceptions (Bongos, A-Zandeh &c.), while the 
Eskimo, who ought apparently to be brachycephalic, are on the 
contrary extremely dolichocephalic. 

To meet the endless transitions between the 
two extremes, Broca has proposed a convenient Doifcho- f 
fivefold division 3 , which being frequently referred Mesati-and 

. ... .... J _ , Brachycephali. 

to in anthropological writings, is here appended : 

1. Dolichocephali, with index No. 75 and under. 

2. Sub-dolichocephali, ,, 75*01 to 7777. 

3. Mesaticephali, ,, 7778 to 80. 

4. Sub-brachycephali, ,, 80 -or to 83-33. 

5. Brachycephali, ,, 83-34 upwards. 

A few examples of each will suffice for a character which, as 
shown, has mainly a sub-varietal value only : 

i. Dolichocephali. 

Kai-Colo (mean) 65* 

Australian 7 r '49 

Eskimo (Greenlander) 7177 

Neanderthal 7 2 (?) 

Hottentot and Bushman 72-42 
Kafir 72*54 

1 Jour. Anthrop. Inst. Nov. 1880, p. 157. 

2 "L'indice horizontal ne caracte'rise pas les groupes primaires de 1'hu- 
manite. Mais il retrouve toute son importance dans la repartition des races 
appartenant a chacun d'eux " (De Quatrefages, op. cit. p. 215). 

* Rev. d' } Anthrop. 1872, p. 385 et seq. 






Dolichocephali (cont.) 

W. African Negro 


Low-Caste, Calcutta 74- 17 



Berber 74*63 

Nile Nubian 


Laugerie Basse 74*^5 

Algerian Arab 



(Lozere), one 75* 



Dolmens N. of Paris 


Anglo-Saxons 76*10 

Guanches (Canaries) 


Polynesians (some) 76-30 

Old Egyptians 


Copts (Modern Egyptians) 76*39 

Ainus (some) 


Basques of Guipuzcoa 77*62 



Chinese 77-60 

3. Mesaticephali. 

Ancient Gauls 


Hawaiians 80-0 

Mexicans (normal) 


Afghans 79 to 80 -o 



Ossetians 8o'o 



Petit-Morin (Marne) and] 

S. Americans (various) 


others from Neolithic \ So'o 

N. Americans 


Caves and dolmens J 



French Basques 


Italians (North) 81*80 

Low Bretons 


Andamanese 81-87 

Mongols (various) 


Finns 82-0 

Turks (various) 


Little Russians 82-3 



Germans (South) 83-0 

5. Brachycephali. 



Burmese 86* 



Armenians 86*5 

Croatian s 


Solutre, one 88-26 



Peruvians 93-0 



Huxley's Mongol 98-21 

Some value has also been attached to the vertical index (high 
and broad), which, when it rises to or exceeds 100, determines 
the so-called hypsistenocephaly characteristic of the Malicolos and 
other Melanesians. 


But of all cranial measurements none is more important 
than that which determines the varying degrees of 
gnathism, that is, the greater or less projection of Gnathlsm - 
the upper jaw, which itself depends on the angle made by the 
whole face with the brain-cap. The more obtuse the angle, the 
greater will be the maxillary projection (frognathism), the more 
vertical the face, the less the projection (orthognathism) \ Hence 
gnathism, which is best seen in profile, is indicated by the so-called 
"facial angle,' 7 accepted by all anthropologists as one of the best 
criteria of race. The evolution, which is intimately associated with 
the dentition and change from raw to cooked food, has obviously 
been from the extreme projection of the higher apes and of primi- 
tive man (see profiles p. 183) to the nearly vertical position of the 
Mongolic and Caucasic groups. Hence prognathism is naturally 
regarded as characteristic of the lower, orthognathism of the higher 
races. " The profile of the face of the Calmack is almost vertical, 
the facial bones being thrown downwards and under the fore part 
of the skull. The profile of the face of the Negro, on the other 
hand, is singularly inclined, the front part of the jaws projecting 
far forward beyond the level of the fore part of the skull. In the 
former case the skull is said to be orthognathous, or straight -jawed ; 
in the latter it is called prognathous --a term which has been ren- 
dered with more force than elegance by the Saxon equivalent 

snouty 2 ." 

Combining this feature with eurygnathism 8 , that is, lateral pro- 
jection of the cheek-bones, GeofTroy Saint-Hilaire found that the 
Caucasic face is oval with vertical jaws; the Mongolic broad 
(eurygnathous); the Negro prognathous; the Hottentot both pro- 
and eurygnathous. 

Nevertheless Topinard, who has made a special study of 
gnathism in all its bearings, distinguishes between a superior and 
an inferior facial angle, the former (general facial gnathism) being 
fallacious, the latter, that is, sub-nasal gnathism, being alone trust- 
worthy. "Anthropologists have been wrong up to the present 
time in giving so much importance to the projection of the whole 

1 Gr. 6p06s, straight ; 7rp6, before ; yvAOot, jaw. 
3 Huxley, Man's Place in Nature, p. 146. 
8 Gr. etfptfy, wide, broad. 




Facial Index. 

maxilla, or of the whole face.... There is no uniformity of results 
in any given series, the most flagrant contradictions 
being met with between averages in allied races.... 
But sub-nasal, or true prognathism, furnishes of itself the dif- 
ferential character of the various human types 1 ." Sub-nasal 
gnathism is determined by the angle formed by a line drawn 
from the nasal spine (sub-nasal point) to the anterior extremity of 
the alveolo-condylean plane. This plane, which gives the total 
projection of the skull, is about parallel with the horizontal line of 
vision, coinciding with a line drawn from the alveolar point 
(median point of the alveolar arch) at right angles to a perpen- 
dicular falling from the occipital condyles. Topinard gives the 
subjoined table of results : 

True or sub-nasal prognathism. 






From this table it appears that the facial is never a right angle, 
so that absolute orthognathism does not exist. All races are more 
or less prognathous, , the European least, the Negro most, the 
Mongol and Polynesian intermediate. In Europe the most ortho- 
gnathous appear to have been the Gauls, Corsicans and Neolithic 
men, the Finns the least. The high position of the Tasmanians in 
the series is remarkable and puzzling, one of those disturbing 
elements that render all classifications so hazardous. Otherwise 

Individual extremes 

89 to 5 1 *3 Merovingians 

8 (White races 

8 2 76-5 

Finns and Esthonians 


76' ,, 68-5 


< (Black 

69 ., 59'5 











Dead Man's Cave 


New Caledonians 






W. African Negroes 



Namaquas and Bush 


1 Anthropology t Part n. ch. iii. 




After von Bauer. 



the difference between Caucasian and Mongol is very marked, 
while from the latter to the Negro the transition is gradual. " The 
Negroes of the east coast of Africa are less prognathous than those 
of the west ; the Negroes of Oceania less than those of Africa ; the 
purest Hottentots reach the highest maximum of the whole human 
race " (ib. p. 282). 

The above-mentioned correlation of the teeth with gnathism 

gives to this character a racial value scarcely if at 

The Den- ^ inferior to that of the facial angle itself. Of the 


facts already determined the subjoined are amongst 
the most important. 

Sir W. Flower shows that the molars are larger in the lower 
races, where they may occupy on the alveolar arch the same com- 
pass as in the chimpanzee. That this relation has persisted from 
the remotest times is evident from the fact that in the man of Spy 
(p. 146) "the molars increase in size posteriorly to the same extent 
that they do in the apes, which is the reverse of what is usual in 
man, where they diminish posteriorly, or in a few lower races 
(Australians &c.), remain equal 1 ." In this palaeolithic race the 
premolars approximate "the relative dimensions seen in the 
chimpanzee," while the third molar even exceeds that of the 
chimpanzee, "reminding one of some of the gibbons" (ib. p. 333). 
Thus may perhaps be explained the curious fact that, as noted by 
Dr Houze, " the third molar is often as large as the others in the 
lower races V' whereas in Europeans the last molar is disappearing 
through disuse, so that the jaws contract and prognathism dimin- 
ishes, as already shown by Darwin and Mantegazza. To this 
contraction, however, is due the marked irregularity of the dentition 
in civilised man, the teeth getting crowded together for want of 
space in the shrunken jaws. In savage tribes this defect scarcely 
occurs, but on the contrary supplementary teeth appear, as amongst 
the New Caledonians, where Bertillon and Fontan have noticed a 
fourth molar. "Cette anomalie est en rapport avec le caractere 
infe'rieur du prognathisme et Fampleur de la machoire 3 ." 

1 E. D. Cope, The Genealogy of Man, in The American Naturalist, April 
1893, P. 333- 

- Bui. d. 1. Soc. d'Anthrop. de Bruxelles* 1894, p. 136. 

3 Houze*, ib. p. 137. In the same place Bourgade is referred to as stating 


On the other hand wear and tear depends mainly on the 
quality of the food, hence varies with the social conditions, being 
greater amongst the lower (coarse eaters), than amongst the cul- 
tured classes. Consequently, however paradoxical it may sound, 
the intelligence would appear to be in inverse ratio to dental wear ; 
" plus celle-ci [usure] est considerable, plus celle-la [rintelligence] 
est rudimentaire " (/#.). In a word it must be obvious that use 
and disuse necessarily play a vast part in the character of the 
masticatory apparatus, which is otherwise so persistent, and con- 
sequently such a valuable test of race. Every morsel of food taken 
into the mouth at once brings the teeth and jaws into play; hence 
these organs, remaining for ages unchanged in unchanged sur- 
roundings, may be modified with relative rapidity by change of diet 
through altered habits of life. 

Few physical characters yield more uniform results than does 
the nose, which is normally thin, prominent, long, 

. , . , , , , v . . The Nose. 

straight or else convex (arched or hooked) in the 
higher races, in the lower short, broad, more or less concave and 
even flat. A careful study of this organ shows almost better than 
any other the coordination of parts in the facial features gene- 
rally. Thus the small flat concave is usually correlated with high 
cheek-bones and narrow oblique eyes (Mongol); the short with 
wide nostrils and depressed root, with everted lips and bombed 
frontal bone (Negro); the short with blunt rounded base and 
depressed root, with heavy superciliary ridges and long upper lip 
(primitive Australian and Tasmanian); the large, straight or arched, 
with regular oval features (Semite and European). Hence the nasal 
index, which expresses the relation of the maximum breadth of the 
anterior orifice to the maximum length from the nasal spine to the 
naso-frontal suture, is regarded by Broca as one of the best tests 
of ethnical differences. Note that the nasal spine, or sub-nasal 
point, lies at the base of the outer or lower extremity of the carti- 

that amongst the New Caledonians the canines "tres souvent de"passent en 
longueur le niveau des autres dents." It is further pointed out that the jaws of 
the pariah dogs in Constantinople are wolfish, while the tenderly nurtured 
King Charles has lost the typical dentition of the species. " Au lieu d'avoir 
six molaires et pre"molaires au maxillaire superieur ils n'ont plus que trois ou 
quaere, et les cuspides sont pour ainsi dire nulles " (F. L. de Pauw, ib. p. 140). 




laginous septum separating the nostrils ; also that the centre of the 
naso-frontal suture (where the nasal joins the frontal bone), lies at 
the root of the nose midway between the orbits (sockets of the 
eyes). Taking the length as 100, the relation varies 
absolutely from 72*22 in a Bushman to 3571 in a 
Russian, and between these extremes are distinguished three 
groups, as under : 

Nasal Index. 

1. Platyrrhinian\ with wide nasal skeleton (all 

Negroid races; most Mongols) . 

2. Afesorrhinian 1 , intermediate (all Americans except 

Eskimo) ....... 

3. Leptorrhinian l , elongated (Caucasic races; Es- 

kimo) ........ 

A few examples in ascending order will suffice : 

Nubian Negroes 
W. African Negroes 
New Caledonians 




Basques (French) 
Basques (Spanish) 





49' 2 5 


447 1 

In the eye both colour and shape have to be considered. 
The colour of iris and sclerotic is of less value in 

Colour and . . . 

shape of the the higher than m the lower races, where it is more 
cye - uniform, more persistent and more generally cor- 

related to the complexion. Thus the European iris is of every 
shade from black to brown, hazel, and light blue, although even 
here dark is normal in the Melanochroid division, light in the 
Xanthochroid ; sclerotic in both whitish. But in the Negroid 
and Mongoloid groups the iris is generally almost black, or deep 

1 Gr. irXemJj, broad, flat; pfoos, middle, median; \6irr6j, slender, thin; 
fa (Gen. /5n>6s), nose; terms introduced by GeofTroy Saint-Hilaire to classify 
the monkey tribe, and later applied to the Hominidse. 


brown, the outer being always darker than the inner circle; sclerotic 
of the Negro yellowish. The shape has no great racial value, 
except in the Mongol division, where it is characteristically slant, 
with outer angle turned upwards, and the inner often covered 
by a fold of loose integument. This occurs even amongst some 
Eskimo, covering the caruncula lachrymalis, and " forming, as it 
were, a third eyelid in the form of a crescent 1 ." The Semitic 
eye is also somewhat almond shaped, or at least more oval than 
the European ; but the character is not constant. 

Sometimes the face as a whole is mentioned as of distinctive 
value ; but this is rather the result of diversely com- 
bined elements than an additional factor. According 
to the form and disposition of the orbits, forehead, 
nose, cheek-bones, jaws, lips, &c., the features assume a general 
expression, a racial physiognomy, which is sufficiently constant, 
though liable to be affected by dress and ornament. The average 
observer notices, not so much particular points, as this general 
expression of the countenance, which was often correctly repro- 
duced by the Egyptian, Babylonian, and Assyrian artists. Thus 
have been transmitted from early historic times several racial types 
Akkad, Semite, Hittite, Hamite, Negro. There are broadly dis- 
tinguished four characteristic faces : 

Simian, due to extreme prognathism, seen best in profile : 
Negro; Negrito. 

Broad and flat, due to lateral projection of cheek-bones and 
small nose, seen best in front ; Mongol. 

Hatchet-shaped, due to lateral projection of the maxillaries; 
Prairie Indian. 

Regular, determined by orthognathism, oval contour, large 
nose, small mouth, straight eyes ; Caucasic races. 

Stature, like the eye, is more uniform amongst the lower than 
amongst the higher races, where it is largely affected 

,.-... Stature. 

by pursuits, town or country life, agricultural or in- 
dustrial occupations in mines or factories, and so on. Hence 
there are not only tall and short Americans, such as the Patagonians 

1 King, Physical Characters of the Esquimaux, in Jour. Ethnol. Soc. 
Vol. I. 1848. 




and their Fuegian neighbours, but also tall and short Englishmen, 
and even tall and short Londoners, as is evident by comparing the 
East End population with those of "club-land." 

Excluding the abnormal dwarfish and gigantic specimens of 
the showmen, the height ranges from about 1-40 to i'8o metre 
with a mean of 170, or, say, between 4 feet 7 inches and 6 feet 
2 inches with a mean of 5 J feet ; this for the male adult, from 
which for the female must be deducted about 8 per cent, in the 
tall and 5 per cent, in the short races. The sub- 
joined tables of heights, chiefly from Broca and 
Topinard, show that all the Negritoes are dwarfish, 
the true Negroes tall, the Mongols rather below the average, the 
Americans extremely variable 1 : 

Tables of 

Tall Races: 1*70 upwards. 

Australians (some) 1718 

Scandinavians i -7 1 3 

Scotch 1710 

English 1708 

Eskimo (Western) I 73 

170 to 1-65. 

Eskimo (Central) ^654 

Caucasus tribes (some) 1*650 

French 1*650 

Hindus and Dravidians 1*645 

Jews 1-637 

Magyars 1-631 

Nicobar Islanders 1-631 

Chinese 1-630 
Araucanians & Botocudos 1-620 

Sicilians 1-618 

Finns 1-617 

Indo-Chinese 1*615 

1 The figures are in metre with 3 decimals as allowing greater accuracy than 

vulgar fractions, to which they may be roughly reduced by making i metre 
in. and '05 = 2 in. Thus, Wissmann's Batwa: 

1 '40 = 39$+*6= 551=4 feet 7^ in. 



Brown Polynesians 




W. African Negroes 




Middle sized : 







Charuas (S. America) 

i -680 

Arabs (some) 


Germans (some) 


New Caledonians 


Fuegians (some) 







J ' 6 57 




Short : i *6o and under. 


i -600 

Melanesians (some) 






Australians (Sydney) 




Orissa tribes 




Kurumba (Nilghiri) 


Batwa (WissmannV) 




Wambutti (Stanley's 2 ) 


Akka (Emin's 8 ) 


In general the stature, as applied to all the Hominidae, would 
appear to be a question of averages, almost more than any other 
important physical character. The absolute range, however, as 
here seen, is limited to about two feet (Patagonians Wambutti). 

In recent years anthropologists have made systematic studies of 
several other anatomical points, such as size of the 
pelvic basin, relative length of the extremities, span ca ?factors hysi " 
of the outstretched arms, finger markings (Galton). 
Some of these have doubtless some racial value, and when applied 
to a sufficiently large number of subjects from various peoples may 
be expected to yield good results. But most of the points vary 
too much to be of any service in determining human varieties, 
though useful in identifying individuals 4 ; hence their increasing 
interest in connection with the new science of " Criminal Anthro- 
pology," cultivated especially in Italy and France. 

1 My Second Journey through Equatorial Africa, 1891, p. 365. 

2 In Darkest Africa, 1890, II. p. 150. At p. 92 Stanley states generally 
that the Wambutti "vary in height from three feet to four feet six inches*'; 
but the above is the shortest measured by him. Three feet, or a little over, 
have also been spoken of by other travellers ; but no trustworthy measurements 
of adults seem to fall much below about 4 feet. 

3 "The measurement of their height I have taken from Emin Pasha's 
anthropological notes ; he has measured a good number of them, mostly women ; 
but men or women have never exceeded 4 feet i inch in height" (A. J. 
Mounteney-Jephson, Emin Pasha and the Rebellion at the Equator, 1890, 

P- 372). 

4 " The other parts of the skeleton also have differences more or less pro- 
found in the different ethnic groups the stature, the length of the extremities 
both absolutely and relatively to the stature and to the trunk, the thoracic 
form, and so on. But such differences are but slightly characteristic in com- 
parison to those presented by the brain -case and the face " (Sergi, Le Varieta 
Umane, quoted in Nature, April 18, 1895, p. 595. 



Cranial Capacity Size of brain and Mental Capacity correlated in the animal 
series and partly in man Comparative Tables of Cranial Capacity 
Language the chief mental criterion Relation of speech to Anthropology 
Phonesis a physical function which cannot be neglected by the anthro- 
pologist Value of language to the ethnologist Evolution of speech 
from the inorganic to the organic state The faculty originated most pro- 
bably in a single centre Reply to the linguistic polygenists Speech of 
relatively recent growth Hence at first unstable and subject to great 
fluctuations Hence also linguistic divergence more rapid than physical 
types, forming species and genera which cannot mix Hence no mixed 
languages Consequent value of speech as a racial test Linguistic more 
easily distinguished than physical groups Table of mixed peoples speak- 
ing unmixed languages Table of peoples whose speech has shifted with- 
out mixing Table of peoples whose physical type has changed, their 
speech persisting Hence speechand race not convertible terms But speech 
often a great aid in determining ethnical elements The morphological 
orders of speech Old views of linguistic growth The "Root" theory 
Monosyllabism not the first but the last stage of growth The sentence 
the starting-point The monosyllabic languages originally polysyllabic 
Chinese the result of phonetic decay The Aryan root theory exploded 
Root and Atom; Sentence and Molecule Agglutination Its nature and 
test The morphological orders not fixed species but transitional phases 
of growth Inflection reverts to Agglutination Agglutination passes into 
Inflection and Polysynthesis Polysynthesis not a primitive but a late 
condition of speech Differs in kind from Agglutination Nature of In- 
flection Diagram of linguistic evolution Development of speech not 
linear but in parallel lines Synthesis and Polysynthesis tend towards 
monosyllabic analysis Change from pre- to post-position in the Aryan 
group Change the Universal Law of all living speech Social state : Fish- 
ing, Hunting, Agriculture, no test of race Social Usages poor criteria 
Religion Origin and development of nature and ancestry worship 
Anthropomorphism due to the common psychic character of man Hence 
common religious ideas no proof of common origin or of contact Like 
usages no evidence of common descent. 

A? already remarked, the size as distinct from the shape of the 

Cranial skull, gives its volume or "capacity," which although 

capacity. to k e care f u iiy distinguished from the mental capacity 

(p. 42), stands, nevertheless, in close association with the mental 

characters. As the size of the brain-pan is necessarily correlated 

to the volume of its contents, the brain; so on this volume to some 


extent depends the quality of the mind, of which the brain is the 
organ. The lighter and smaller the organ that is, smaller 
relatively to the whole organism the weaker, cczteris paribus^ will 
be its functional power. Hence the cranial capacity, although a 
physical factor, serves as a connecting link between 
the physical and mental criteria ; and if gradation and Mental ' 

can here be shown between different races, we shall 
be able to speak on solid grounds of high and low 
varieties of the Hominidse. The limitations of each will also be 
more clearly seen, and the inherent inequality of the various 
members of the human family made evident against the precon- 
ceived theories of sentimentalists. On this basis, for instance, it 
might be fairly argued that man, specifically one on the physical 
side, may not be so on the mental. In the lower orders of the 
animal series the gradation in question undoubtedly exists, the 
ratio between weight of brain. and body diminishing rapidly in the 
ascending order, thus : 

Fishes i to 5,668. 
Reptiles i to 1,321. 

Birds i to 212. 
Mammals i to 168. 

Similarly between the highest anthropoid and the lowest human 
brain there is a tremendous gap, that of the gorilla weighing only 
20 oz. while that of the most degraded savage weighs 32 oz. in a 
body scarcely half the weight of the gorilla's. Again, according to 
Morton the size of the smallest human skull, as measured by its 
capacity, is 55*3 cubic inches, that of the largest 114, while the 
difference between the smallest and that of the gorilla is consider- 
.ably more than that between the smallest and largest normal 
human brains 1 . Herbert Spencer considers the brain of civilised 
man nearly 30 per cent, larger than that of the savage. 

But, as explained in Chap. III., mental power depends also 
on the number of cerebral convolutions, and still more on the 
quantity of grey cortical substance contained in both hemispheres. 
Here also the gulf between the lower and higher orders is vast, 
though relatively slight between the anthropoids and man. " Be- 
tween the smooth brain of the wistitis (lowest of the Hapalidae), 
and the marvellously complicated brain of chimpanzee and orang, 

1 Charles Bray, Anthropology ', p. 23. 




there is a gap, while there are but faint shadows of difference 
between the latter and that of man. The enormous and complex 
mass of convolutions in man... is composed of the same funda- 
mental folds, united by the same connections and separated by the 
same sulci V Nevertheless, as already pointed out, considerable 
differences exist between the human varieties in 
respect of the secondary convolutions, which are 
smaller and more complex in the higher races. But 
this subject has hitherto been little studied, and 
ethnology has still to depend mainly on comparative tables of 
cranial capacity (volume), such as those here appended from 
Topinard and Barnard Davis. 




English and Scotch 1427 

Germans 1382 

Austrians 1342 

French 1334 

African Negroes 1238 

Annamese 1233 

Cape Negro 974 

Barnard Davis. 




Dahomn n 


Subjoined is Morton's table, re-arranged by de Quatrefages, 
showing mean capacity in cubic inches : 

English 96 

Germans 90 

Anglo-Americans 90 

Arabs 89 

Grgeco-Egyptians 88 

Irish 87 

Malays 86 

Persians 84 

Armenians 84 

Circassians 84 

Iroquois 84 

Lenap^ 84 

Cherokees 84 

Shoshons 84 

African Negroes 83 

Polynesians 83 

Chinese 82 

Hindus 80 
Egyptians (Ancient) 80 

Fellahs 80 

Mexicans 79 

Peruvians 75 

Australians 75 

Hottentots 75 

1 Broca, Memoire sur les Primates, 1869. See also Topinard's Diagram, 
p. 40. 


The incongruities of this table have already been noted, 
and such tables have as a rule but little value, the observations 
being seldom made on a sufficient number of specimens. 

A better index of the difference between the mental capacity 
of the various human groups is afforded by the 

r i^ r i u s.- i A. i_ ^ Language 

reasoning faculty, of which articulate speech is at the chief 
once the measure and the outward expression. But crtaSon. 
this special characteristic of man, in virtue of which 
he stands entirely apart from and immeasurably above all other 
creatures, has hitherto had the misfortune of suffering from friend 
and foe alike ; over-zealous philologists ranking it much too high, 
anthropologists depreciating it to a corresponding extent. Thus 
while the latter too often decline to recognise its claim to con- 
sideration in ethnological studies, many of the former go so far as 
to assert with Horatio Hale that language is the true Relation 
basis of anthropology, that by their speech alone of speech to 

. .. /. , -r- if i -/- j Anthropology. 

the tribes of men can be scientifically classified, 
their affiliations determined, and their mental qualities discerned ; 
hence the logical inference that linguistic anthropology is "the 
only Science of Man 1 ." 

But apart from such extravagant assumptions, even the purely 
anthropological student must recognise the importance of this 
faculty, when it is pointed out that different phonetic systems 
imply greater or less differences in the anatomical structure of the 
vocal organs. 

Owing to such differences Europeans, for instance, find it 
impossible, after years of residence amongst the ph neg . 
natives, to pronounce the various clicks of the Bush- a physical 
man, Hottentot and Kafir tongues, the splutterings 
and other harsh sounds of the Thlinkft, Chintik (not the jargon). 
Apache and some other American idioms, the gutturals ( o, , .. 
?* J>) f * ne Semitic group, and so on. The " absolute impossi- 
bility " of imitating certain utterances in some of the New Guinea 
languages is by Miklukho Maclay rightly attributed to "funda- 
mental differences in the anatomical structure of the larynx and 

1 Language as a Test of Mental Capacity -, in Transactions of the Royal 
Society of Canada* 1891. 

K. I 3 


the whole muscular system of the organs of speech in the two 
races 1 ' [European and Papuan]; he adds that "not only the 
organ of speech but also that of hearing plays an important part, 
for the same ^word may be heard in a totally different manner by 
different individuals 1 ." The new school of phonetics associated 
with the names of Ellis, Bell, Sweet, Jespersen, Paris and others, 
is entitled to look for aid on these points from special anthro- 
pologists, who on this ground alone cannot afford to neglect 
linguistic studies. They should be able to tell us, why the planta- 
tion negroes, whose mother tongues have for several generations 
been English, French, Spanish or Portuguese, still continue to 
speak these languages barbarously, why the same languages 
continue to be a " shibboleth " to the Jews resident for hundreds 
of years amid the European populations, and why no Ephraimite 
could frame to pronounce this very word right, whence 

" so many died, 

Without reprieve adjudg'd to death 
For want of well pronouncing shibboleth." 

Samson Agonist es. 

But if the anthropologist has no time to take heed of these 
things, he should at least understand that they possess no slight 
racial value. Phonesis, now recognised as the true basis of all 
philological studies, "belongs almost exclusively to the physio- 
logical characters of race 2 ." 

In any case the evolutionist, who regards articulate speech as 

Value of lan- a natura ^ phenomenon, will not hesitate to recognise 

guage to the its value in ethnological studies. The faculty itself, 

ethnologist. , TT . . , , . 

proper to all the Hommidse, and to them only, 
would alone suffice to separate them as a distinct family from the 
other anthropoids. As soon as the term alalus drops out of 
HaeckePs definition of man's precursor, we get the homo primi- 
genius himself, the origin of the human race being coincident with 

1 Ethnologische Bemerkungen uber Papuas der Maclay fCtiste, quoted by 
J. C. Gallon, Nature, Jan. i, 1880. 

2 T. de Lacouperie, Academy, Sept. 4, 1886, p. 156. Prof. H. Schuchardt 
goes further, and boldly asserts that "there is no more difference between 
biology and philology than between biology and chemistry " (Literatnrblatt fur 
Ger. u. Rom. Philologie, 1892); this also on physiological grounds. 


tiat of speech. " If it is language that constitutes man, then our 
.rst progenitors were not real human beings, and did not become 
uch till language was formed in virtue of the development of the 
rain and of the organs of speech " (Schleicher). 

For the evolutionist, who necessarily traces man back to a 
peechless precursor, speech is a function which perfects itself 
iand in hand with the growth of the organ (p. 41). Hence the 
iculty starts from a germ, and its history is one of continuous 
pward evolution from slowly accumulating crude utterances, 
luch utterances, vague at first in sound and sense, 
re to be regarded as the imperfect expression of Sp ^ech! ti0n f 
nward emotion and feelings, of outward things and 
ctions, differing from the accompanying gesture-language only in 
his, that the one appeals to the sense of vision, the other to that of 
learing. Primitive man, always a social being congregating in 
imily groups, expressed his thoughts by speech and gesture, and 
s speech expanded with the infinite capabilities of the vocal 
rgans, gesture fell more and more into abeyance, now surviving 
nly amongst the lower and some of the more emotional higher 
aces (American aborigines, Neapolitans). 

The first utterances, like those of the higher apes, were doubt- 
ess mere jabberings 1 , scarcely more distinct and varied than the 
>resent language of man's companion, the dog, who, a howler in 
he wild state, has learnt in domesticity to bark diversely, to 
ell, yelp, growl, snarl, whine, whimper, moan, or bay 2 . So with 

1 Prof. R. L. Garner's recent experiments have not convinced us that 
'monkeys talk, the power of expression being commensurate with that of 
bought." This is the error into which Max Mliller and other Hegelians have 
ullen. Thought (reason) and language are not convertible terms, and it is 
onceivable that even the homo alalus might have arrived at a considerable 
egree of culture by the aid of gesture language, ejaculations, an upright 
'osition and specialised hands and feet. In any case a varying range of thought 
annot be denied to the speechless ape and other dumb creatures. " Man does 
ot speak because he thinks. He speaks because the mouth and larynx com- 
lunicate with the third frontal convolution of the brain. This material con- 
.ection is the immediate cause of articulate speech " (Andr< Lefevre, Race mid 
'Miiguage, 1894, p. 3). 

2 " La domestication supprime les inquietudes de la faim, et donne des 
Msirs au systeme nerveux. L'id&ition augmente, et s'exteriorise par une 
lodification de 1'appareil phoneteur" (Houze, loc. cit. p. 135). 



our pliocene precursor's cry, which is rightly regarded by Andre 
Lefevre as "the undoubted embryo of speech 1 ." With the growing 
needs of society it could not fail to develop by various processes 
mimesis, reduplication, repetition, stress, prolongation of vocalic 
sound 2 sufficient raw material for the constitution of the inorganic 
or first phase of human speech, which may well have been reached 
in eolithic times. At least the Tasmanian, practically at the 
eolithic stage of culture, spoke a tolerably developed language, 
which had fully passed beyond the inorganic 3 to the early organic* 

That this marvellous evolution occurred more than once in a 
few independent centres is conceivable, but improbable, and it has 
been seen in Chap. vu. that the existence of radically distinct stock 
languages is no argument for a multiple origin of human speech. 
All the conditions seem best accounted for by the assumption of a 
single centre of evolution, coincident in time and place with the 
evolution of man himself. The faculty once acquired would thus 
have accompanied man in all his migrations over the globe; it 
was never lost, and all members of the family, however debased, 
are found in full possession of this priceless heirloom. Had the 
faculty risen independently in several centres, this need not, pro- 
bably would not, have been the case. Some tribes, migrating 
from the common centre to unfavourable regions, or surrounded 
by unfavourable conditions of existence, might well have remained 

1 Op.cit. p. 22. 

2 Rival schools have advocated now one now another of these processes, 
which have thus been brought into ridicule and stigmatised as the "bow-wow," 
the "pooh-pooh" or other theories. But all have in varying proportions con- 
tributed towards the formation of language, and some (reduplication) have 
entered into the very structure of the highest forms of speech. The past tenses 
of all our English "strong verbs " are due to reduplication. 

3 These terms are here used merely in the sense of coherent^ incoherent; 
organised, unorganised^ and not as used by A. W. Schlegel and the host of 
German metaphysical philologists, who, despite Pott's protests, still persist in 
speaking of language as, not merely figuratively, but actually an organism, a 
concrete substance existing, growing, flourishing and decaying independently 
of the human organism. Articulate speech should rather be likened to the 
notes emitted by musical instruments of varying degrees of perfection. Heyse 
(quoted by Sayce) calls language "the music of the soul," though in a sense 
different from that here implied. 


speechless, or the faculty might have been arrested at various low 
stages of the inorganic phase. But no such speechless or semi- 
speechless tribes have ever been discovered, so that the universal 
diffusion of the faculty is itself an argument in favour of its dis- 
persion from a single centre. Or let us suppose hundreds of 
speechless groups scattered over the primaeval woodlands. The 
odds are that some of them will remain in that state, for evolu- 
tion, even granting the conditio sine qua non, is not a necessity. 
Nowhere are the conditions more favourable for wheat-growing 
than in California ; yet not an ear ever ripened in that region till 
the seed was planted by the discoverers. Moglichkeit, say the 
Germans, ist nicht Notwendigkeit. 

But it may be argued that the alaloi may have existed, but 
were either killed off by the speaking tribes, better 

J o? Reply to the 

equipped for the struggle, or else learned to speak linguistic 
from them. But if killed off, we are not concerned Pol ye enists - 
with them, any more than with the Homo alalus himself, common 
precursor of the assumed speechless and speaking tribes. That 
they could not have learnt to speak is obvious from the fact that 
the faculty, as explained, is of slow growth, its development going 
on simultaneously with that of the vocal organs. 

Again it may be urged that languages differ specifically and even 
generically from one another; hence must have had independent 
centres of origin. This point has been referred to in Chap, vn.; 
but as it is the source of endless misconceptions in ethnology, it 
will be desirable here to dispose of it once for all. In his assumed 
speechless precursor man has physically a real starting point. 
From that precursor he ascends directly, and owing to the per- 
sistence of physical characters has, relatively speaking, diverged 
little from that prototype. But with language the case is entirely 
different. Its starting-point was not, and could not be a fully 
developed prototype, but only a germ, that is, such 

Spcecn of 

inarticulate utterances as may have been inherited relatively 
from the precursor. Physically man goes back recent e rowth 
through imperceptible transitions to the lower animal series ; 
linguistically he goes back no farther than the Homo alalus. Hence 
speech, as compared with physique, is an entirely new feature of 
relatively recent growth. Now all new features are at first incon- 


stant, pliable, unstable, until permanently fixed. Compare the 
tendency of new varieties of the pigeon or geranium to sport and 
revert. Renan well remarks that " linguistic families apparently 
isolated could have had fruitful contacts at times when they were 
still capable of being re-cast. In speaking of languages, we can- 
not too carefully distinguish the embryonic state, during which 
accidents harmless in riper age may have had vital consequences, 
from the perfect state, when they are fixed, as it were, in a definite 
mould 1 ." 

It was during this " embryonic state," which is here called the 
inorganic phase, that, as the stuttering groups spread 
unstable! 1 ^ 5 * abroad from a common centre, their speech, such 
as it was, rapidly diverged, and broke readily into 
numerous varieties. Then these varieties, following each its 
inward bent, gradually acquired greater consistency and firmness. 
They grew from varieties into species and genera, while the speakers 
have continued to remain mere physical varieties to the present 
time. Hence it is that within the same physical group, the Caucasic, 
for instance, we find several linguistic groups differing generically 
(Aryan, Semitic, Georgian, Chechenz, Basque &c.). This pheno- 
menon, which has been the cause of such wonder and of so many 
delusions, thus appears simple enough, and indeed inevitable, 
when we but reflect that, cateris paribus, linguistic change far more 
rapidly than physical types. But these changes have been in 
progress since pleistocene times, when the groups of speaking 
Hominidae were already spread over the face of the globe. Con- 
sequently the divergence is now too great to trace the linguistic 
groups back to their primordial inorganic condition. A primordial 
organic condition could never have existed for all the Hominidse, 
who carried with them from the centre of dispersion nothing but 
a common stock of incoherent utterances. It follows that a 
common language of organic type is a chimaera which will always 
elude the grasp of linguistic monogenists. 

Hence also another curious result, of paramount importance in 
the study of linguistic and ethnical groups. The human groups, 
being mere varieties, all amalgamate with each other; but the 

1 VOrigine des langues t p. 212. 


linguistic groups, forming species and genera, never amalgamate, 
but on the contrary are mutually repellent There is no such 
phenomenon as linguistic miscegenation, no change of inner 
structure by any amount of contact, but only word- borrowing, and 
the words so borrowed have all to conform to the genius of the 
languages into which they are accepted. There are many mixed 
races; indeed, as seen, all races are mixed; but there are no mixed 
languages, that is, mixed in the sense here explained. 
The so-called jargons Chintik, Pigeon- English, the 
" Slavo-Deutsches " and " Slavo-Italisches " of Prof. 
Schuchardt 1 , and so on, are not mixed languages, but rather mixed 
vocabularies with little trace of the grammatical forms of the 
idioms from which the words are brought together and gradually 
organised on a fresh basis. An attempt has been made by 
Mr J. C. Clough to prove the existence of mixed languages in 
a "prize essay" which rather proves their non-existence. The 
writer relies mainly on sound and vocabulary, which are not in 
dispute, while his references to grammar are highly uncritical 2 . 
English, if any, might be called a mixed language ; but its grammar 
is purely Teutonic, and while it has embodied thousands of Latin 
and French words, it has not embodied a single Latin or French 
grammatical form. So with Hindustani, Persian, Turkish, and ail 
the other so-called " mixed languages," none of which are mixed 
in their inner mechanism. " Never has the grammatical structure 
of a language accommodated itself to a new one, but rather the 
whole language has disappeared, and has been supplanted by the 
new one; for such a change in the structure of a language would 
presuppose a transformation of ideas and the mode of con- 
necting the elements of thought, which we deem next to im- 
possible 3 ." 

1 Gratz, 1885. 

2 The Existence of Mixed T,anguageSi c., 1876. The uncritical character 
of this essay appears from such statements as that at p. 7, where the Romance 
tongues are stated to have been " once nothing more than jargons of various 
Gothic and Latin dialects'*; that English Grammar "has become Romance in 
spirit " (p. 95) ; and that the transposed Hindustani form/*/// mard ki for mard 
kijdtivs "according to the Persian order" (p. 18). 

8 Waitz, op. cit. p. 248. 


From this " kakogenesis " of speech, taken in connection with 
the "eugenesis" of races, there follow some im- 
vaiueofspeech portant inferences. Arguing from the universal 
as a Racial miscegenation in Hungary, Schwiker concludes that 
"speech remains the most conspicuous distinctive 
indication of European affinities 1 ." De Quatrefages points out 
that " had it not been for their special language no one would 
have hesitated to consider the Basques as belonging to the same 
family as other southern Europeans 2 ." Sayce also observes that 
" the physiological races of the modern world are far more mixed 
than the languages they speak ; the physiologist has much more 
difficulty in distinguishing his races than has the glottologist in dis- 
tinguishing his families of speech 8 ." Thus we have in Europe 
mixed Keltiberian peoples, but no mixed Keltiberian languages; 
Finno-Slavs, Siavo-Teutons, Kelto-Teutons, but no Finno-Slav, 
Slavo-Teutonic or Kelto-Teutonic tongues. The inferior, and 
sometimes even the superior races, have in all cases abandoned 
their mother tongue, while adopting, without seriously modifying, 
that of the conquerors or conquered, as the case may be. Within 
two generations the victorious Northmen of the Seine valley forgot 
their Norse speech and adopted the Romance of their Gallic sub- 
jects. These Gauls themselves had, on the other hand, previously 
changed their old Keltic speech for the Latin of their Roman 
masters. In this region of Northern France there have thus arisen 
racial complexities of all sorts, but never any permanent linguistic 
confusion, one language simply displacing another without pro- 
ducing any hybrid forms of speech, which, if they exist at all, are 
certainly the rarest of philological phenomena. The Basques of 
Navarra are at present slowly giving up their old Escuara tongue 
for Spanish, but they do not blend the two into some new Hispano- 
Basque variety. So with the Pruczi, or "Old Prussians" of 
Lithuanian speech, the nearly extinct Wendish Polabs, the Ugric 
Bulgarians, many Permian Finns, Kelts of Cumberland, Cornwall, 
and Ireland, all or most of whom have been assimilated in speech 
to the surrounding Slav and Teutonic populations. The same 

1 Siatislik des Konigreichs Ungarn, Stuttgart, 1877, p. H& 

2 7'he Human Species, 1879, p. 434. 
8 Science of Language, I. p. 366. 



2O I 

process is at present going on amongst the Finnish Vepses of Lakes 

Onega and Ladoga, who are becoming rapidly Russified in speech 

while retaining their distinctive physical features 1 ." 

This important subject may be further illustrated by 

the subjoined table of some of the chief European 

peoples, showing that all belong ethnically to mixed 

groups, but linguistically to unmixed families : 

Table of 
mixed peoples 
speaking un- 
mixed lan- 


English; Scotch 
Irish (West) 
Russians (many) 

Hungarians (Magyars) 
Prussians (East) 

Ethnical Group. 
Siluro-Kel to-Teutonic 













Here we have no compound terms, Le. mixed elements, in the 
linguistic column, while in the ethnical all the terms are compound. 
Moreover, there is no doubt at all as to the linguistic terms, whereas 
the ethnical are largely conjectural, or merely symbols (Silurian, 
for instance) of unknown elements. It is this consideration that 
to some extent justifies the remark of Waitz that " for the classifi- 
cation of mankind philological research has given much more 
certain and harmonious results than the physical study of man 8 ." 
But a little reflection will show that too blind a trust in philology 
may lead to as erroneous results as too blind a trust in craniology 

1 " Les Vepses disparaissent en prenant la langue russe, mais ils se con- 
servent fort bien au point de vue anthropologique " (Ch. de Ujfalvy, Bui. d. /. 
Soc. de Gtogr. Paris, 1877, p. 320). 

* Anthropologie der Naturvolker^ part I. sect. 5. 




Table of 
peoples whose 
speech has 
shifted with- 
out mixing. 

or in other physical characters. A language, for instance, may in 
the struggle for existence get killed oft", although the people 
speaking it may escape the 'same fate. Countless 
instances may have occurred in past ages, many 
have occurred within the historic period, of such 
linguistic shiftings. Subjoined are a few instances 
of communities which are known to have changed 
their languages in comparatively recent times : 

Present Speech. 



On the other hand cases may arise of the reverse process, that 
is, of peoples gradually changing their physical type, while retain- 
ing their original speech. Such instances would not affect the 
truth of the general statement that physique is more persistent 
than language, because the change would be mainly due to mis- 
cegenation, a most potent factor in modifying physical, but, as 


Original Speech. 



Irish (many) 


East Prussians 











Non-Italic (?) 

Hazaras; Aimaks 


Polabs (most) 


Burgundians, ) 
Franks, Lombards] 


Permian s (many) 


Basques of Vitoria 


Bretons (many) 


Takings (many) 




Samangs (many) 


Griquas (many) 


Bantu 1 

Negroes of America 

^ T , \ &c. 

Negroes of Madagascar 



seen, powerless to modify linguistic types. The change in 
question would be likely to occur when conquering or intruding 
races find themselves strong enough to maintain 
their political and social supremacy, but numerically Table of 
too weak to prevent fusion and assimilation with 

the subject peoples. Subjoined are some of the has changed, 

J . . r ,.,, .., their speech 

most conspicuous instances, for which there is either persisting. 
linguistic or direct historical evidence : 

Race. Original Type. Present Type. . . S ^ ch -. 

* Jtr Jl (unchanged). 

Western Turks Mongolic Caucasic Mongolic 

Magyars Mongolic Caucasic Mongolic 

Finns (many) Mongolic Caucasic Mongolic 

Basques Non-Aryan Aryan Non-Aryan 

Berbers (many) Hamitic Semitic Hamitic 

Abyssinians (some) Semitic Negroid Semitic 

Tibus (some) Hamitic Negroid Hamitic (?) 

Germans of) _, . ~ . -, ^ . 

_ \ Teutonic Georgian Teutonic. 

Caucasus J 

The last instance is most remarkable, and well deserves the 
consideration of those anthropologists who attach but little im- 
portance to the influence of the environment, and still less to 
the value of speech as an aid to the ethnologist. The Germans 
in question, a few hundred Wurtembergers, who settled (1816) at 
Yelisavethpol near Tiflis, had originally fair or red hair, light or 
blue eyes and broad coarse features. In the first generation brown 
hair and black eyes began to appear, in the second black eyes and 
hair became the rule, while the face acquired a noble oval form, 
and these changes were due entirely to the surroundings, no 
instance of crossings with Georgian natives being on record. At 
the same time these transformed Wurtemburgers continue to speak 
their German mother-tongue uninfluenced by the local dialects 1 . 

1 II paratt que dans 1'espace de deux generations les colons suabes ont 
change* physiquement d'une maniere remarquable sous I'mfluence du milieu. 
Quoiqu'il n'y ait point eu de croisement entre eux et leurs voisins, la plupart 
ont maintenant la chevelure fonc^e, les yeux noirs, la figure ovale et reguliere, 
la taille elegante et souple. Us ne ressemblent plus a leurs cousins rested dans 
la mere-patrie " (Reclus, vi. p. 225). 


From these tables, establishing an apparent antagonism 
between speech and race, it follows that, as Prof, 
speech and Sayce rightly remarks, " philology and ethnology 
vertiwV terms are not convertible terms 1 ." But this writer goes 
too far, much too far, when he adds that " identity 
or relationship of language can prove nothing more than social 
con tact... Language is an aid to the historian, not to the ethno- 
logist... Language in short was not created until the several 
types of race had been fully fixed and determined. The 
xanthochroid and the melanochroid, the white albino and the 
American copperskin existed with their features already fixed and 
enduring before the first community evolved the infantile language 
of mankind " (id.). Prof. Max Miiller cannot approve of this 
view, because it summarily disposes of his contention that speech 
and reason are one. It is not to be supposed that groups of 
speechless Hominidse could become highly specialised as Homo 
Caucasicus ("xanthochroid and melanochroid"), Homo Amcricanus 
&c., without a liberal endowment of reason 2 , which would thus 
have existed for ages without the faculty of speech. 

But the view must be rejected on other grounds, and the last 

highly gratuitous assertion has in fact already been disposed of 

(pp. 196-7). The statement that language proves 

But speech SO cial contact only and is no aid to the ethnologist, 
often a great ^ o > 

aid in deter- implies a fundamental misconception of the correla- 

mining ethni- r . _ . 

cai elements. tlon f speech to race. Cases may and do arise, 
where language will infallibly prove the presence 
of distinct ethnical elements, which, but for it, would never 
have even been suspected, much less determined. In Europe a 
case in point are the Basques, shown by their speech to be at 
least partly descended from a pre-Aryan or a non-Aryan race, 
which has elsewhere apparently disappeared, but which has far 
more probably become amalgamated with the intruding Aryan 
peoples. Thus from the Basque language we learn to be cautious 

1 Science of Language, II. p. 317. 

3 It would need, for instance, a considerable degree of intelligence for the 
speechless successors of Homo neanderthalensis (still of somewhat Simian type) 
to build the neolithic monuments of Mauritania, Brittany and Britain, or even 
to fashion the delicate palseoliths of the Solutrian period. Ethnological specu- 
lation cannot be safely indulged in from the subjective standpoint. 


in speaking of the peoples of West Europe as of "pure Aryan 
stock." Moreover, if the late G. von der Gabelenz be right, the 
Basque language, connected by him with the Berber, supplies the 
clue to the identification of the non- Aryan element, which would 
thus appear to be African Hamite 1 . So with the Finns, whose 
Uralo- Altaic speech reveals the Mongolic ethnical element, which 
could otherwise be only suspected in their physical constitution ; 
and so with the Magyars, in whom but for their Finno-Tatar 
idiom that element would not be suspected at all, so "Aryanised" 
are they. A Malay element in the Negroid peoples of Mada- 
gascar is placed beyond doubt by their Malayo-Polynesian 
dialects. Or are we to suppose that, by crossing from the 
mainland to the neighbouring island, the Mozambique Bantus 
forgot their mother-tongue, and began to speak Malay somehow 
wafted with the trade-winds from Malaysia across the Indian 
Ocean to Madagascar? Language used with judgment is thus 
seen to be a great aid to the ethnologist in determining racial 
affinities, and in solving many anthropological difficulties. 

It would even appear that the great divisions of speech corre- 
spond, at least to a limited extent, with the great Themor ho 
divisions of mankind. Languages are by most logical orders 
philologists grouped according to their morpho- speec 
logical structure in four main classes or orders, which are distri- 
buted amongst the main branches of the human family as 
under : 

( Most Negroid peoples. 

AGGLUTINATING: J A11 Mongols, Tatars, and Finns. 

Malays and Polynesians. 
\^ Some Caucasic peoples. 
POLYS YNTHETIC : Most American Aborigines. 
INFLECTING : Most Caucasic peoples. 
ISOLATING : Indo-Chinese and Tibetans. 

Here it will be noticed that the last three answer fairly well to 
so many distinct sections of mankind, while the first alone com- 
prises several different groups, the reason being because aggluti- 
nation itself is not of one kind, as is often supposed, but of many 

1 Die Veiivandtschaft des Baskischen mit den Berbtrsprachen Nord-Afrikas 
nachgewiesm, Brunswick, 1894. 


types. All four belong to the organic phase of speech, from which 
it follows that all known tongues, having in prehistoric times passed 
through the inorganic state, have, so to say, already completed 
their natural evolution, at least to the same extent that the 
Hominidse themselves have completed their natural evolution. It 
is commonly assumed by philologists that for language this 
evolution consists of passing successively through the lower to the 
higher of the above specified morphological states, the isolating 
and inflecting being taken as respectively the two extremes, in 
the ascending order, of this assumed linguistic gamut. Opinions 
vary greatly as to how the complete process is accomplished, and 
owing to the difficulties involved in the application of this cut and 
tj . dry theory, its correctness has in recent years been 

Old views j j i J 

of linguistic seriously questioned. But most philologists have 
growt ' till recently accepted the views of Grimm and 

Schleicher, according to which the inflecting state was reached 
through the agglutinating from the isolating, this last consisting of 
detached "roots," such as those yielded by the analysis of the 
Aryan tongues. These roots being monosyllables, it was naturally 
inferred that the Indo-Chinese family, consisting also mainly of 
monosyllables, must represent the most primitive condition of 
Monosyiia- human speech. Thus were established two fallacies, 

not the one that primaeval speech consisted of a mono- 

first but the J * 

last stage of syllabic root-language, the other that Indo-Chinese 

growth. ig still - n that primitive state. The first fallacy is 

now exploded, and it is generally understood that monosyllabism 
is not a necessary condition of primitive speech, Sayce amongst 
others holding that, on the contrary, the sentence was the necessary 
starting-point. Rightly understood, this position is impregnable, 
for the object of speech must always have been to communicate 
thought, and the definition of the sentence is, not a number of 
abstractions called roots thrown together anyhow, but a number of 
terms so arranged as to convey a concept, to communicate thought. 
But this may be done by a single ejaculation such as hush ! or 
whist! and the first sentences could not have been much more 
complicated than such utterances, the full meaning being, where 
necessary, eked out with the aid of gesture language, as still in 
Grebo (West Coast of Africa) and some other savage tongues. 


"This complex of sound and gesture... was the earliest sentence 1 ," 
in which the question of monosyllabism does not enter. 

The second fallacy still persists, and most philologists continue 
to regard the Indo-Chinese languages as in a 
primitive state of monosyllabism from which they The mono- 

1 J J syllabic Ian- 

have never emerged. Yet their original polysyl- guages origi- 
labic character, already dimly seen by Edkins and {UJjj* polysyl ~ 
de Rosny, has now been placed beyond doubt by 
the researches of the late Terrien de Lacouperie. Edkins was 
able to trace /#, great, back to a fuller form dap; yi, one, to tit; 
tsie through tsit to tsik, a joint, and so on 8 . But de Lacouperie 
went further, and recovered not merely monosyllabic but trisyllabic 
forms, such as tadaka, to doubt, now worn down to i*. This view 
is accepted by Mr Robert K. Douglas, one of the first living Sino- 
logists, who writes : " I quite agree with the opinion expressed by 
the late Dr T. de Lacouperie, that the present monosyllabism of 
Chinese, instead of being evidence in support of the theory that in 
their earliest stage all languages were monosyllabic; is another 
proof of the existence of phonetic decay in this 

r r J Chinese the 

and in other tongues. When we trace back Chinese result of Pho- 
to its earliest recognised form in Akkadian, we see netlc ecay * 
unmistakable evidence of the same kind of phonetic decay. Thus : 

Akkadian. Chinese. English. 

Gush-kin Kin Gold 

Ukush Kut, Kwa Gourd 

Kur-fi Ki A fowl 

Inim Nien, Nim To recite 

Garshan Shan Mountain 

Billudu Lut, Lii A statute 

Guk-kal Ku A sheep 

Guk-kud Kit, Kie A wether 

Ukkin Kien All 

Dim-menna Men, Wen Inscription 4 ." 

1 Sayce, op. cit. J. p. 116. 

2 Introduction to the Study of the Chinese Characters, 1876. 

8 Note communicated to A. H. Keane, and published in his Asia (Stanford 
Series), 1882, p. 700. 

4 MS. note communicated to A. H. Keane, Oct. 29, 1894. The Akkadian 
words supplied by the Rev. C. J. Ball. 


But the same process had already made considerable ravages 
thousands of years ago in Akkadian itself, in which the monosyllable 
g$, for instance, has been traced by Mr T. G. Pinches to twelve 
originally distinct words, such as get, root, gtn, seed, gig, night, 
geme, like, gi-num, fire, gin, shekel, &, one 1 . Similar tendencies 
are at work in other groups, such as Tibetan, Danish, English 
(most monosyllabic of all Aryan tongues), Otomi of Mexico, and 
especially Yoruba, Tshi, Ewe, and other allied idioms of Upper 
Guinea, where, as in Indo-China, the numerous homophones re- 
presenting originally distinct terms are now distinguished by their 
different tones. 

Monosyllabism is thus shown to be, not the first but the last 
stage in the evolution of human speech, and the 

The Aryan r 

Root Theory numerous theories based by Bopp, Schleicher and 
expio e . others on the assumed original monosyllabic state 

of the " Aryan roots " all fall to the ground. All the facts tend to 
the conclusion that primitive speech was not monosyllabic or 
isolated, but on the contrary involved, after it had passed the 
inorganic phase, and it may be regarded as certain that at no time 
did man ever speak in "monosyllabic roots." A root is a pure 
abstraction, the residuum of a term stripped of its formative 
elements, comparable to the atom of physicists, not exactly a 
fiction, but an ultimate particle of matter, which eludes the 
keenest analysis, and which has no independent existence in the 
cosmos. But as the atom unites with one or more atoms to form 
the molecule, so the root unites with one or more roots to form the 
sentence, the unit of speech. The combination, however, is much 
closer in the physical than in the linguistic order ; hence the che- 
mical union of parts in the molecule is represented by the much 
looser agglutinative process, which is thus seen to constitute the 
first stage in the organic condition of speech. Roots, therefore, 
whether mono- or polysyllabic, must be relegated with the atom to 
the ante-cosmos, and agglutination of some form taken as the 
primary condition, the first morphological state of all organic 
languages, which either remain in that state, or else pass on to 
the three other above specified morphological orders. 

1 Observations on the Early Languages of Mesopotamia, paper read before 
the R Asiat. Soc., March 17, 1886. 


The character of the several orders is determined by the way in 
which the relational combine with the notional . . 

, . , . . j i i Agglutination. 

elements. The relational were themselves doubtless 
originally notional, as it is almost impossible to imagine the 
deliberate invention, except perhaps later by analogy, of meaning- 
less particles, such as the Turkish mak, far, mi, me, or the Aryan 
in, un, ab, ex, introduced for the purpose of expressing in combi- 
nation the various relations of the notional words. But these 
particles were unsuited to enter into true combination until they 
had gradually ceased to become notional, and until they 
had at the same time been reduced by phonetic decay to con- 
venient adaptive forms. When they are merely tacked or glued, 
as it were, on to the notional words, language enters the aggluti- 
nating, i.e. the first strictly morphological state, of which there are 
divers kinds and grades, according to the various ways and 
degrees of combination. But, in general, the true test of agglu- 
tination is the power of the particles to become detached and 
shift their places in the combined form, as when ly in the English 
word manly makes room for ful in man-ful-ly\ so the Turkish 
sev-mek, sev-il-mek, sev-il-me-mek, &c. ; and the Assamese manuh- 
bilak-or, of-the-men (plural bilak inserted between noun manuh 
and its gen. case-ending or). A vast number of languages are of 
this agglutinating order, from which all the others have emerged in 
diverse directions 1 , although this evolution of speech has been 
denied by Sayce and some other reactionists against the old 
theory of root origin. Thus Sayce speaks of "the magical frontier 
between flection and agglutination," which can never be "cleared," 
" since to pass from agglutination to inflexion is to revolutionise 
the whole system of thought and language and the 
basis on which it rests, and to break with the past lo JcaidT 
psychological history and tendencies of a speech," transitional 
(ib. i. p. 131). Nevertheless this break with the 
past has been made, and as Prof. Jespersen shrewdly 
remarks, "revolutions do take place in the world of languages, 

1 " So far as verified facts in linguistic history go, all outward devices of 
derivation and accidence grew out of agglutination, that is, by adding originally 
independent [notional] words" (G. v. d. Gabelentz, die Sprackwissensckaft, 
1891, p. 189). 

K. 14 


even if they take more time than it takes the French to change 
their constitutions. If a thousand years suffice to change a type 
of speech like that of King Alfred into the totally different one of 
Queen Victoria, then the much longer period which palaeontologists 
and zoologists accord to mankind on this earth could work still 
greater wonders.... Sayce stands, with regard to these three or four 
types of speech, in much the same attitude which naturalists kept 
with regard to the notion of * species' before Darwin came 1 ." It 
is argued that the transition from a significative term to a forma- 
tive element is an unknown, or at least an extremely rare 
phenomenon, because but few particles in current languages have 
been traced back to notional words, so that most of them must 
be accepted as ' ' meaningless affixes " from the first *. Not so ! This 
is rather a case in which it may safely be argued from the few to 
the many, for the process, so far from being a "rare phenomenon," 

inflection * s norma ^ m most languages, though arrested by 
reverts to various causes in cultivated idioms. The above 

gg u ma ion. exam pj es f rom English and Assamese show that 
reversions may take place from inflection to agglutination, which 
in fact is a general tendency amongst the Gaurian (Neo-Sanskritic) 
tongues of India, and also to a less extent in Italian and other Neo- 
Latin tongues. Thus Italian incorporates both direct and indirect 
pronominal object, as in da(m)-mi-lo = give-to-me-it (sing.); date- 
me-lo - give-to-me-it (plur.); dando-me-lo = giving-to-me-it (pres. 
part.). In the same way the whole of the Hindi conjugation 
except a solitary tense (the so-called " aorist ") has become par- 
ticipial with gender and number but no person^ as in so many 

1 Progress in Language, 1894, p. 132. 

2 According to Ludwig's "adaptation theory," as soon as the relations of 
words to each other in the sentence got to be understood, "pre-existing suf- 
fixes," no doubt floating about in circumambient space, were set apart to 
determine them. Thus the Greeks captured the suffix es, which in ir6d-c<r-fft 
and iro8-fo-wv (irodwv) has no grammatical meaning, but which came to symbo- 
lise the nom. pi. in ir6$cs and the 2nd pers. singular in trvires, being thus made 
to do duty for the plural in nouns and the singular in verbs. Thus also we are 
back in the old days, when speech was regarded as an elaboration of the 
conscious will, instead of being the result of unconscious cerebration acting 
through the vocal organs, as it had previously acted through the facial muscles, 
say, in the miocene precursor. 


agglutinating systems. Similarly vernacular Bengali is now mainly 
agglutinating, forming nominal cases, number, and gender by 
juxtaposed nouns, the case-endings themselves being the same for 
singular and plural, while conjugation is effected by verbal nouns, 
or participles and auxiliaries. "In a word the whole language 
tends to become reduced to nouns, joined together to express 
declension and conjugation 1 ." What is happening so generally 
during the process of disintegration (synthesis to analysis) must 
have taken place universally during the process of integration in 
the pre-inflecting agglutinative stage of linguistic evolution. 

From that stage language developed, according to its different 
initial tendencies, in various directions towards 
complete decomposition, as in the above-described tj^* 8 ^^*" 
isolating state of the Indo-Chinese group ; partial into inflection 

, . . . . . , , - , and Polysyn- 

decomposition, as m the particle languages of the thesis. . 
Malayo-Polynesian group; polysynthesis, as in 
most of the American groups \ and synthesis, as in the inflecting 
Aryan, Semitic and Hamitic groups. Polysynthesis, regarded by 
some philologists as a primitive condition of speech, is on the 
contrary the outcome of great phonetic corruption, syncope and 
clipping of words and particles, which become so fused together 
that the sentence often tends to assume the aspect of a single 
composite term. Its involved character has no doubt been 
greatly exaggerated by Duponceau*, who introduced the word 
"polysynthetic," describing it as a process by which the greatest 
number of ideas are comprised in the least number of words. 
But the fact remains that in Iroquoian and other 
languages of this type "the stem of a verb or adjec- 
tive may be combined with the stem of a noun 3 " ; Agglutination 
and thus arises practically unlimited participial con- 
jugation, which is the essentially distinguishing feature of polysyn- 
thesis, as compared with all other incorporating systems. Basque, 

1 Ch. Johnston, Paper read at the Oriental Congress, Sept. 1891. See also 
Asiatic Quarterly Review for July 1892. 

- Memoire sur le systeme grammatical de quelques nations indiennes de 
FAmtrique du Nord, 1838, and in other writings. 

8 J. N. B. Hewitt, Polysynthesis in the Languages of the American Indians, 
in The American Anthropologist, Oct. 1893, p. 387. 



for instance, the most highly agglutinating tongue in the Old 
World, may run the changes through such a form as I-giving-it-to- 
//, stopping short at the direct and indirect pronominal objects; 
but Iroquoian introduces the noun as well, and inflects I-gvving-it-the- 
book-to-him \ and as the pronominal elements are few, the nominal 
innumerable, it is at once seen that agglutination and polysynthesis 
differ not merely in degree, as is often assumed, but also in kind. 
Fusion by syncope, however, which is only one aspect of poly- 
synthesis, is a common phenomenon, as seen in the English 
ha'porth - half-penny-worth ; the Mexican teo-calle - teotl-calli = 
God's'-House (temple); the Vei (W. Coast of Africa) nkumbafdwuye 
= n-kumu m-be a fo wii-ye = I-tell-you-this ; the Basque arkume = 
ardi-hume sheep-little = lamb \ the Spanish hidalgo = hijo-de-algo = 
son-of-somebody = noble; the French (as pronounced) kekcfcxa = 
qu'est-ce que c'est que cela, and so on. But what is more or less 
exceptional elsewhere is normal in polysynthesis, which is thus 
seen to represent a very advanced state of development, in fact 
standing in this respect on the same level as inflection. In the 
Nature of inflecting (synthetic) order, as represented by the 
inflection. Aryan group, the so-called "monosyllabic root" 
is assumed to develop a " stem " (Lat. am, am-a), with which are 
fused one, two or more relational particles, so as to form one 
inseparable compound, such as the Lat. am-a-b-u-nt-u-r, where the 
cohesion is complete, and not loose, as in Turkish (see above). 

Taking the more or less articulate cry as the starting-point, the 
various morphological orders of speech would thus appear to have 
been evolved from a primitive inorganic condition, 
through various types of agglutination, to poly- 
evolution. synthesis, inflection and isolation, as in the sub- 
joined diagram : 



Inorganic ^^Agglutination 

W Jfe -^ 


This genesis of speech explains the reason why the later forms 


possess some features common to the earlier stages through which 

they have passed. They resemble the higher orders in the 
Animal kingdom, whose embryonic life is an epitome of their past 
history. But in the Animal kingdom the evolution was not 
through all the four classes of fishes, reptiles, birds and mammals, 
but either from fishes through reptiles to birds, or from fishes 
through reptiles to mammals. So in the linguistic kingdom, the 
evolution was not through all four morphological orders, in 
single linear direction, but from the earlier agglutinating phases 
to inflection, isolation, and polysyn thesis in parallel lines as 
above. Thus linguistic growth must be admitted, not only 
within each order, which Sayce allows, but also from order 
to order, which Sayce denies. If the evolution be eternally 
within each order, and if the transition from one order to another 
be impossible, then the orders themselves must be conceived as 
having come into existence independently as we now find them. 
For the evolutionist this would be like saying that the animal 
classes came into existence ready made, say, by creative force. 
The very difficulty which often presents itself of drawing a hard and 
fast line between the several orders is itself a proof that, servants 
servandts, they may pass by imperceptible transitions one into the 
other. Thus Finnish and Tamil have developed agglutinating 
forms which are scarcely to be distinguished from true inflection, 
while Basque holds a somewhat intermediate position between 
agglutination, inflection and polysynthesis. It is the " ornitho- 
rhynchus " of the linguistic family, and to express this anomalous 
position W. von Humboldt classed it by itself as 
einvfrleibend, " incorporating." Hence these orders 
must be regarded as progressive phases, not as fixed 

linguistic species. And if it be objected that some Monosyllabic 
languages have never got beyond the agglutinating a y sis - 
state, the answer is that some animals have never got beyond the 
classes of fishes, or reptiles. But as they were subject to perpetual 
change in the past, so they are never at rest in the present. 
English amongst the inflecting tongues of the Old World has 
made vast strides towards monosyllabic isolation, as seen in such 
an expression as " town talk," where a noun becomes an adjective 
and a verb a noun without any change of form, but, as in Chinese, 


by position alone. So in the New World polysynthesis is no 
longer universal, and Otomi amongst other formerly polysynthetic 
tongues has by constant wear and tear at last simulated the 
appearance of a monosyllabic isolating language 1 , although one of 
its dialects (Mazahua) is still " decidedly polysynthetic 3 ." 

But inflection may also tend towards true polysynthesis, as in 
Sanskrit, where the limits are not always observed 
from'pre- to between the word and the sentence. Thus in 
positions virtue of the euphonic rules of Sandhi a group of 

words may flow together, developing such a form as 
trinairgunatwamapannairbadhyante^ rivalling in length the formid- 
able compounds of Cherokee, Mexican and other American 
incorporating tongues. Hence the inevitable revolt against such 
monstrosities, which has resulted in the Neo-Sanskritic postfixing 
vernaculars. Thus the change from the prefixing to the postfixing 
principle, which some hold to be impossible, has actually taken 
place during the historic period within the Aryan group itself, just 
as in prehistoric times prepositions would appear to have been 
preceded by postpositions. Thus, such forms as me-cum, te-cum, 
se-cum, rare in Latin, are normal in the sister Umbrian of the 
Eugubine tables, where we have such constructions as tertiam-a 
spantim = tertiam-ad libationem ; ocre-per Fisiu = colle-pro Fisio^ 
showing how the post- may have easily become prefixes and then 
separable prepositions 8 ; so tuta-per Ikuvina = civitate-pro Iguvina 
(Eugubium, Gubbio) ; fratus-per Attiiedies=fratribus-pro Attidiis ; 
asam-ad ad aram ; spiniam-ad = ad mensam; uvi-kum = cum ove; 
esunek esunu anter = inter istud sacrificium, &c. Compare also the 
Latin urbem versus with the Hinddstdni shahr-ki-taraf-(meu) = city- 

1 " L'Othomi nous a tout Tair d'une langue primitivement incorporante, et 
qui, parvenue au dernier degre* d'usure et de de'labrement, a fini par prendre 
les allures d'une dialecte &. juxtaposition " (Charancey, Melanges de Philologie et 
de Pattographie, 1883, p. 80). 

2 A. S. Gatschet, MS. note communicated to A. H. Keane, Nov. 14, 1894. 
This great authority is also inclined to remove Kwakiutl, Ata'kapa, Isleta and 
others of the Tehua group, as well as the Chibcha of S. America, from the 
polysynthetic order. "The Chibcha is remarkably simple; it approaches 
monosyllabism, and shows no incorporation" (MS. note, Dec. 4, 1887). 

8 "Nous croyons que 1'usage des postpositions a precede celui des preposi- 
tions" (Michel Bre"al, Les Tables Eugiibines % 1875). 


of -direct ion (in), where nouns are transformed to postpositions. 
When it is remembered that analogous transformations took place 
in the agglutinating Akkadian of Babylonia, it will 
be seen that inflection must not be rated too highly universal 

above agglutination, of which it is "merely a child 1 ." }* 

Thus the transition from order to order is established 

for all periods and for all linguistic groups, and there is no natural 

division between the historic and prehistoric life of languages, as 

maintained by the Hegelian school of philology. Change, the 

universal law, is arrested only by the extinction of species. 

In his progressive development as a social being man passes 
necessarily from the hunting and fishing to the pastoral 
and agricultural states. But these higher states are in 
all cases determined, not by race but by the outward conditions. In 
Africa the Negro is normally a husbandman, although in all other 
respects greatly inferior to his Arab neighbour, who, as a rule, is a 
herdsman. In the Upper Lena basin the Yakut domesticates the 
horse, in the Middle the reindeer, in the Lower the dog. The Tungus 
of North-east Siberia tills the land in the fertile Amtir basin, tends 
the herd farther north, hunts and fishes in still higher latitudes. 
The Arab is a nomad pastor on the steppe lands of Nejd, a good 
agriculturist in the rich, well-watered upper valleys of Yemen. In 
a word, peoples pass so obviously from one to another of these 
states, that they can in no way be accepted as distinctive character- 
istics of any race. They are the proper subject of ethnography 
and geography, rather than of ethnology in the stricter sense. 
Hence they are discarded by Dr E. Hahn, who substitutes six 
" Kulturformen " distributed, irrespective of race, throughout so 
many geographical areas over the surface of the globe 8 , all existing 
side by side, and determined by the physical and climatic condi- 
tions of the several regions. 

1 " Ueberhaupt darf man die Flexion, die, abgesehen von dem innern 
Lautwandel, nur ein Kind der Agglutination ist, nicht zu hoch, und die Ag- 
glutination nicht zu niedrig anschlagen " (Dr C. A. F. Mahn, Denkmdler der 
baskischen Sprache> 1857, p. xxiii). 

2 Petermanrfs Mitteilungen, January, 1893. 


The same remark also applies in great measure to the primi- 
tive condition of the family and tribe: to such 

Usages poor ... . , . , ,, 

Criteria of institutions as polyandna, polygamy, monogamy, the 
race - matriarchal and patriarchal states, exogamy, en- 

dogamy, the totem systems, tattooing, cannibalism, and similar 
practices more or less common to all primitive communities. The 
investigation of such subjects, however interesting in itself, can 
throw little light on the origin and mutual relations of the funda- 
mental divisions of the human family. They belong to the natural 
as distinguished from the recorded history of man, and come more 
specially within the province of the historian of the growth of 
human culture. W. Earl 1 remarks that in Malaysia the grades of 
civilisation depend rather on the physical conditions than on race. 
Near the sea and rivers the people become fishers and navigators, 
on the uplands tillers of the land, and so on, and the observation 
may be taken as of general application. 

So also with the various religious systems of mankind, even the 
most primitive of which betray evidence of growth 

Religion. fc _ ... ... . _ . 

from some still more primitive previous state. It is 
obvious that, apart from the question of direct revelation, with 
which we are not here concerned, all natural religions must have 
had their beginnings in the first faint awakenings of the reflective 
powers. As soon as man began to remember his dreams, and to 
take cognisance of himself in a dim way as something distinct 
from his surroundings, all natural phenomena must have presented 
themselves to him as the effects of causes beyond his control and 
comprehension. With the growth of the reasoning faculties, 
comparisons would be instinctively made between such phenomena 
and those dependent on his own will. Thus the human powers 
and passions became the standards to which all things were 
referred, and instead of man being fashioned to the likeness of his 
deities, his deities or demons were rather fashioned to the likeness 
of man 2 . Hence the good and evil spirits take the complexion of 
the times, reflect the social status of the community; to their 

1 Native Races of the Indian Archipelago >, p. 235. 

2 Hence J. P. Richter's remark that "minder der Mensch nach Gottes 
Bilde geschaffen sei, als dass er sich seinen Gott nach seinem Bilde zu schahen 
pflege " (quoted by Cams, op. cit. p. 94). 


friendly or hostile feelings are attributed all favourable or adverse 
events. All nature is filled with such invisible agencies, which 
move about freely, as man moves freely in his dreams, and which 
have to be propitiated or enlisted in his service by offerings and 
other devices. Some are the spirits of the mountain, the forest, 
the storm or the flood, some the spirits of departed men 
themselves; whence nature and ancestry worship. Such must 
have been the beginning of the anthropomorphism 
which is the essence of all primitive, and of many 
later religions. 

Whether this anthropomorphic state was reached character of 
before or after the first dispersion of the human man * 
family could matter little. The conditions being everywhere 
alike or analogous, the evolution of this early phase of natural 
religion must have everywhere proceeded along the same lines of 
thought. Whether developed in a common centre, or in several 
independent centres, the religious sentiment would still present 
but slight shades of difference, such as might arise, for instance, 
from the differences between the manifestations of the natural 
forces in hot, cold, dry or moist climates, in high or low latitudes, 
in mountainous or forest regions. The sun, naturally regarded as 
a supreme agent in tropical lands, might be replaced by the moon 
under more temperate climes, and it is noteworthy that the 
gender of sun and moon, respectively masculine and 
feminine in the south of Europe (sol, luna) is 
reversed in the north (A. S. sunne, mbna). In the no proof of 

... . . . . . common 

same way night becomes the measure of time with origin, 
the Teutonic, day with the Italic peoples. But 
apart from such easily explained discrepancies, the early religions, 
growing out of a common anthropomorphism into all shades of 
fetishism and shamanism, would everywhere present substantially 
the same general features \ hence could nowhere serve as distinc- 
tive marks of the primary human groups. 

It is further to be noticed that religious ideas, like social usages, 
are easily transmitted from tribe to tribe, from race to race. 
Hence resemblances in this order, where they arise, must rank 
very low as ethnical tests. If not the product of a common 
cerebral structure, they can prove little beyond social contact in 


remote or later times. A case in point is the remarkable parallelism 
between the four great scenes of the Buddhist purgatory depicted 
on the Japanese temple scrolls, and the corresponding scenes on 
the road to spirit-land depicted in the Aztec Vatican codex: 

Buddhist Purgatory. 

1. Soul wades across the 
river of death. 

2. Passes between two iron 
mountains pushed together by 

3. Climbs mountains of 
knives which cut its hands and 

4. Is gashed by knives fly- 
ing through the air. 

Aztec journey to Spirit-land. 

1. Soul crosses the river of 

2. Passes between two 
mountains that clash together. 

3. Climbs a mountain set 
with obsidian knives. 

4. Is beset by these knives 
blown about by the winds. 

The parallelism is complete; but the range of thought is 
extremely limited nothing but mountains and knives, besides the 
river of death common to Egyptians, Greeks and all peoples 
endowed with a little imagination. Hence Prof. E. B. Tylor, who 
calls attention to the points of resemblance 1 , builds far too much 
on them when he adduces them as convincing evidence of pre- 
Columbian culture in America taking shape under Asiatic in- 
fluences. In the same place he refers to Humboldt's argument 
based on the similarity of calendars and of mythical catastrophes. 
But the "mythical catastrophes," floods and the like, have long 
been discounted, while the Mexican calendar, despite the authority 
of Humboldt's name, presents no resemblance whatsoever to those 
of the " Tibetan and Tartar tribes," or to any of the other Asiatic 
calendars with which it has been compared. "There is absolutely 
no similarity between the Tibetan calendar and the primitive form 
of the American," which "was not intended as a year-count, but as a 
ritual and formulary," and whose signs " had nothing to do with 
the signs of the zodiac, as had all those of the Tibetan and Tartar 
calendars 8 ." Regarding all such analogies as may exist " between 

1 Mythical Beliefs as Evidence in the History of Culture. Paper read at 
the British Association, Oxford, 1894. 

9 D. G. Brinton, On Various supposed Relations between the American and 


the culture and customs of Mexico and those of China, Cambodia, 
Assyria, Chaldaea and Asia Minor," Dr Brinton asks pertinently, 
" Are we therefore to transport all these ancient peoples, or repre- 
sentatives of them, into Mexico?" (ib. p. 147). So Lefevre, who 
regards as "quite chimerical" the attempts made to trace such 
resemblances to the Old World. " If there are coincidences, they 
are fortuitous, or they result from evolution, which leads all the 
human groups through the same stages and by the same steps 1 ." 

Many far more inexplicable coincidences than any of those 
here referred to occur in different regions, where not even contact 
can be suspected. Such is the strange custom of the Couvade, 
which is found to prevail amongst peoples so widely separated as 
the Basques and the Guiana Indians, who could never have either 
directly or indirectly in any way influenced each other. Of these 
Guiana Indians Reclus remarks that, to whatever ethnical group 
they may belong, their customs are everywhere very much alike : 
" L'analogie du milieu et des conditions e'conomiques Like 
a rapproche les populations 8 ." Sometimes wide- no evidence of 
spread customs which appear motiveless, and there- scent? " 
fore all the more inexplicable when found prevailing 
amongst distant peoples, may receive quite a simple explanation 
from some circumstance still surviving amongst one or two 
primitive peoples. Thus the strange reluctance of the mother- 
in-law to meet her son-in-law, observed amongst Papuans, Austra- 
lians, Zulus and some American aborigines, seems accounted for 
by a Patagonian practice which persisted till quite recent times. 
On the death of any young person the head of the family was 
required to despatch some aged woman, a mother-in-law by pre- 
ference. Hence through fear of such a fate women acquired the 
habit of avoiding all contact with their sons-in-law, and the feeling 
continued after the motive had been forgotten. Thus the most 
startling coincidences go for nothing, and, speaking generally, like 
usages may be regarded as the least trustworthy of all evidences of 
common descent. 

Asian Races> from Memoirs of the International Congress of Anthropology* 
Chicago, p. 148. 

1 Race and Language > p. 185, 

* Nouvelle Geogr. Universelle, xix. p. 46. 




Four Primary Groups Homo ^Ethiopicus, Mongolicus, Americanus, Caucasi- 
cus Family Tree of the Hominidae The primary groups derived, not 
one from the other, but independently from a common precursor Their 
differences determined by their different environments Position of the 
several groups The Negro The Mongol and American The Caucasian 
Remarks on this Terminology Comparative Table of the physical and 
mental characters of the four primary groups Centre of Evolution Dis- 
tribution of land and water in Secondary and Tertiary Times The Indo- 
African Continent The Austral Continent The Eurafrican Continent 
The Euramerican Continent America accessible from Europe and from 
Asia Theory of de Quatrefages on the migrations of primitive man His 
linguistic argument Views of Dallas and Brinton Evolution "with a 
jump " The Missing Link Probable centre of Evolution and Dispersion 
the Indo- African and Austral regions, true Home of the Lemurs and of 
the Anthropoids Characters of the pliocene precursor and of the pleisto- 
cene sub-groups persistent in the Afro-Austral regions Pliocene and 
pleistocene migrations from the primeval home Order of Development 
of the primary groups in their several centres of evolution Monogenist 
and Polygenist views reconciled Flower and Lydekker on the spread of 
the Hoininidse over the globe. 

WE have seen (p. 166) that recent systematists show a tendency 
to return to the broad groupings of Linne and Blu- 
menbach, and to recognise with them not more than 
three or four main divisions of the human family. 
Flower and Lydekker in a careful survey of the whole field 1 
reduce the Hominidse to three primary groups, the Ethiopia 
Mongolic and Caucasic> leaving the position of the American 
aborigines an open question. Although "inclined to include 
them as aberrant members of the Mongolian type," they add: "It 
is however quite open to anyone adopting the Negro, Mongolian 

1 Introduction to the Study of Mammals > p. 743 et seq* 


and Caucasian groups as primary divisions to place the Americans 
apart as a fourth" (p. 752). But they really go farther than this, 
remarking that "now that the high antiquity of man in America 
perhaps as high as that which he has in Europe has been 
discovered, the puzzling problem, from which part of the Old 
World the people of America have sprung, has lost its significance. 
It is indeed quite as likely that the people of Asia may have been 
derived from America as the reverse 1 ." And as the pre- 
Columbians were practically isolated from the rest of the world, 
" it is difficult to look upon the anomalous and special characters 
of the American people as the effects of crossing a consideration 
which gives more weight to the view of treating them as a distinct 
primary division" (p. 753). It would therefore seem that on 
physical grounds alone these anthropologists are prepared to 
admit the claim of the Americans to be regarded as an indepen- 
dent branch of the human family. This view is greatly 
strengthened by a consideration of the mental characters as 
revealed in the independent cultures of the New World. Hence 
without denying a common origin of both groups, it may still be 
argued that the American offshoot has diverged sufficiently to be 
regarded as a distinct variety in the same sense that the Mongol 
is itself taken as a distinct variety. In other words, the pre- 
Columbians differ perhaps as much from the Asiatic Mongols 
as these do from the Caucasic Europeans ; consequently all stand 
in any scheme on much the same level, constituting three branches 
more nearly akin to each other than any of them is to the Negro 
or Ethiopic branch. 

Linnets original fourfold division (p. 164) must therefore be 
upheld; nothing would be gained either in clear- 
ness or accurac y by attempting on minor considera- 
tions to increase or reduce the number of the great 

Americanus, .,-.,- ,m 

Caucasicus. systematist s primary groups. 1 hese groups, as ex- 

plained, are to be regarded as so many main varieties 

of a single species, and not as so many distinct species of genus 

1 So A. H. Keane : " The arguments brought forward in support of an 
Asiatic origin of the American would not lose their point if adduced in favour 
of an American origin of the Asiatic people" (American Indians, Encylopadia 
Britannica, 9th ed.). 


Homo. As the Negro stands somewhat apart, and admittedly at 
a lower grade than the other three branches, it might be conceived 
as diverging first in the process of upward development from the 
main stem, from which the others ramified later. In a genealogical 
tree of the human family the several primary branches HOMO 
CAUCASICUS would therefore stand relatively to each other and 
to their respective chief sub-branches somewhat as in our Family 
Tree (p. 224). But there is really neither "first" nor "second" 
in the process, and the evolution is to be conceived, not as taking 
place in temporal sequence, but rather as going on all along the 
line simultaneously in space and time. 

What follows will be mainly occupied with an elucidation of 
these genealogies, with a view to the establishment 
of such a general scheme of classification as may rouVderivlid 
seem best to accord with the known pre-historic and from a com- 

, . T . . . . A mon precursor 

present relations. In most ethnological treatises a independently, 
direct transition is assumed from one to another of 
the primary groups, usually from the Negro to the Mongol, 
and from the Mongol on the one hand to the American, on the 
other to the Caucasian. Such direct transitions are in the abstract 
possible, the differences being nowhere more than varietal. But 
they are not probable, and they could scarcely have taken place 
within the limited time available for the implied physical and 
mental changes. For it is to be remembered that the differences 
are shown by comparative craniological studies to have already been 
everywhere established in neolithic times, while the varying grades 
of culture, following in ascending order, show that even in the 
palaeolithic age Europe at least was inhabited by peoples differing 
enormously in mental capacity, consequently it may be assumed 
also in physical appearance. These disparities, presenting them- 
selves at such an early period in the natural history of man, can 
be explained only by supposing them to be the result of develop- 
ments occurring, not consecutively in one area, but simultaneously 
in several areas, and introduced into Western Europe by successive 
waves of migration during the warm inter-glacial epoch. On this 
assumption sufficient time is obtained, not to transform a Negro 
to a Mongol, or a Mongol to a White, which need never have 





happened, but to transform several semi-simian pleistocene 
precursors inhabiting different environments into generalised 
Negro, Mongolo-American and Caucasian precursors respectively 
and independently. Transitions taking place in 
this way would also be immeasurably less violent encea deter"-'" 
than those commonly assumed, as is obvious ; for mined by their 
we need but suppose that the several pleistocene environments, 
groups, already presenting certain differences 
amongst themselves, continued their natural evolution in the 
direction of those differences in varying physical and climatic 
surroundings warm, temperate or cold zones, mountainous or 
low-lying tracts, wooded or open lands, marine or inland regions, 
and so on. Thus under torpid suns it would be advantageous to 
retain more of the original furry coat (hirsute Caucasic Ainus) 
than in torrid lands (Negroes hairless except on head where a 
thick woolly covering is needed). So with a temperate foggy 
climate, which is conducive to a florid complexion, hot suns and 
dry atmosphere which tend to swarthiness, hot suns and moist 
atmosphere which aided by a vegetable diet cause a darkening of 
the subcutaneous pigment. Thus the general evolution would 
appear to have preceded, not in a single or linear direction, but 
as shown in our " Family Tree," by successive lateral rami- 
fications from the parent stem, just as man himself was seen 
(p. 19) to have been evolved not from any specialised anthropoid 
forms, but from the common anthropoid stem by divergence 
antecedent to such specialisations. 

In the Tree the first ramification (to the right) is that of the 
"generalised Negro," that is, the ideal "Homo .. 

*-' ** % Jrosition of 

^Ethiopicus," who during his subsequent natural the several 
evolution largely ceases to be ideal, but retains in * roups * 
more or less modified form a greater or less number of his original 
physical and mental characters. These are never 
entirely obliterated, but continue to flow through 
the arteries of the whole system, as it branches off in various 
directions towards Africa, Oceania and Australia. Hence, despite 
later interminglings, the relationship of the several branches can 
mostly be recognised, thanks to the persistence of a sufficiently 
large number of special features. Where doubt arises, it can only 
K. IS \ 


be through excessive miscegenation, by which the specific unity of 
the whole human family is established, while the more immediate 
kinship of aberrant groups may be obscured. Thus the Australian 
branchlet, pointing towards that of the Toda, suggests possible 
fusion of Melanochroid Caucasic (South Indian) and Austral 
Negro blood at a remote epoch in some now perhaps submerged 
Indo-Austral region. 

After the Negro dispersion the main stem throws off (to the 
left) a generalised Mongolo-American limb, which 
presently breaks into two great divisions, the 
American and the Asiatic Mongol, each preserving 
a share of the common inheritance, but diverging at such an early 
period in their life history that, as above seen, the best authorities 
hesitate with regard to their mutual relationship. The ties have 
been so weakened by long separation that an ideal Homo Ameri- 
canus as well as an ideal Homo Mongolicus must now be assumed 
and studied separately. Here the chief aberrant types, due to 
the same causes as in the Negro family, are the Eskimo, the 
Dravidian and the Finno-Lapp, stumbling-blocks to all sys- 

Between the Negro and the Mongolo-American boughs the 

main stem passes upwards, developing a generalised 

Caucasian. Caucasic type Homo Caucasicus which also at 

an early date ramifies into three great branches, 

filling all the intervening central space, overshadowing the Negro, 

overtopping the Mongol, and shooting still upwards, one might 

say, into almost illimitable space. Such is the dominant position 

of this highest of the Hominidse, which seems alone destined to a 

great future, as it is alone heir to a great past. All the works of man 

worthy of record have, with few or doubtful exceptions, emanated 

from the large and much convoluted brain of the white Homo 

Caucasicus. Needless objection is often made to 

Terminology. . , i i 

this term " Caucasic, which was introduced by 
Blumenbach, and suggested to him by a skull of fine proportions 
belonging to a native of Georgia, South Caucasia. But the word, 
like so many others in scientific nomenclature, is purely conven- 
tional and not restricted to the inhabitants of the Caucasus, who 
are merely taken as somewhat typical members of the whole 


family 1 . It is more important to note that Caucasic is not synony- 
mous with "Aryan," or "Indo-European," as is commonly 
supposed. These are rather linguistic than ethnical designations, 
hence are excluded from our Tree ; and in any case Aryan would 
only form one of many branches of the Caucasic division, which, 
as here seen, also comprises Semites, Hamites, Iranians, besides 
some aberrant groups Tibu, Toda, Ainu, Polynesian and others. 
It follows from the foregoing considerations that by the four 
primary divisions are to be understood those first 
ramifications from the parent stem which, like the Ph /? ic u 1 and 

A 7 mental cnarac- 

branches of a banian tree returning to earth, took tersofthc 

fresh root, and became gradually differentiated gropsf 
independently in so many isolated centres. In 
these centres were evolved those special physical and mental 
characters, the sum of which constitutes the ideal types of the 
several independent groups. And although the ideal types them- 
selves have long ceased to exist in their primordial integrity, the 
determination of the characters is none the less necessary, in 
order to establish distinct standards whereby to fix the position of 
the various sub-varieties in the human family. A comparative 
study of the fundamental types, using the word fundamental not 
absolutely but only in a relative sense, will be facilitated by here 
summing up the more salient features of each division in tabular 
form. These tables, based on the data brought together in 
Chapters VIII. and IX., will enable the student, so to say, to 
reconstruct the ideals by a sort of eclectic process, and thus to 
form some notion of the typical primitive Negro, Mongol, Ameri- 
can and Caucasian, as they may be supposed to have existed in 
their several original homes prior to later migrations and inter- 
minglings : 

1 Hence a distinction might be drawn between the scientific Caucasic and 
the ethnic Cattcasian ; but this is not necessary, as, the explanation once made, 
the sense in which the term is used must always be evident from the context. 
So with the forms Mongol, Mongolian^ Mongolic^ which have similarly a par- 
ticular and a general application, the special Mongol group of Central Asia 
being taken as typical of the whole division. The term Ethiopic has also a 
particular meaning, which sometimes causes confusion. It is applied to the 
eastern branch of the Hamites, who are not Ethiopians in the general sense, but 
members of the Caucasic division. 



cr 5 










Sensuous, indolent, imp 
fitful, passionate an 
though often affectio 
faithful : little self 
hence easy acceptam 
yoke of slavery ; sci 
art uircktvejoped 

Non-theistic, nature 
cestry worship ; fetis 
witchcraft prevalent 

Agglutinating of vario 
and postfix types 

a. Above the average ; 
b. Dwarfish; 4ft. 

Large (macrodont) 

Large, round, pronrinen 
yellowish cornea 


Small, moderately retre 

Prognathous ; 60 

a. Dolichocephalous; ; 
b. Brachycephalous; 8 

a. Blackish; 
b. Yellowish brown 

a. Short, jet black, fri 
in transverse section 
no beard ; 
b. Reddish brown, woo 







3 EL 







< g-.a. 


BS r.'rt U 2 g 





n a 
















.uggish, somewhat sul 
little initiative, but j 
durance ; generally 
thrifty and industri 
moral standard low; 
slightly, art and lette 
rately developed 

olytheistic ; sha 
Buddhism: Transmig 

gglutinating chiefly w 
fixes ; isolating with t 

elow the average ; 5 ft 

[edium (mesodont) 

mall, black, oblique, ou 
slightly elevated, vert 
of skin over inner can 





rominent laterally 

'esognathous ; 68 

rach ycephalous ; 84 



oarse, black, lustrele 
round in transverse 
beardless, but mousta 


w g ^ 3 g" 

3 3 












B s r? 

Moody, taciturn, 
feelings masked 
sive exterior ; in 
physical pain ; sci 
art moderately, I 
ly at all develope 

Polytheistic; anim 

Polysynthetic main 





Medium (mesodont 

Large, bridged 
mesorrhine ; 50 

Moderately prom in 

Mesognathous; 72 

Mesaticephaloub ; 7 

Coppery, yellowish 

Very long, coarse, 
nearly round in se 











R 0* 



P sr <? ^ o. 




(* jr 







9. a > 








.* a 



ctive, enterprising, i 
live ; 
serious, steadfast, so 
stolid ; 
fiery, impulsive, fickl 
science, art and letters 
developed in both 

[onotheism ; Judaism : 
tianity; Mohatnmeda 

hiefly inflecting; some 

Above average ; 5 ft. 
Average ; 5 ft. 5 or 6 

mall (microdont) 

Black ; 
both moderately larj 
always straight 




tnall; unmarked 


Dolichocephalous; 74 
Brachycephalous; 83 



Long, wavy, soft^flas 
Long, straight, wiry, 
both oval in section; b 


ET & 

S' r 

i oo 



% Sg 


! F 







Reserving each of these four types for special treatment, it 
will be convenient here to consider such general questions as 
equally concern them all. Such are mainly their probable prim- 
eval homes, and the direction of their first migratory movements, 
in other words, their several centres of evolution and dispersion. 

From all the foregoing remarks there follows a first important 
corollary, that although man had but one origin, 
one pliocene precursor (p. 38), men had several 
separate places of origin, several pleistocene pre- 
cursors. In our Family Tree four such precursors are assumed, 
and the question at once arises, in what inhabitable regions of the 
globe were they evolved ? Here the inquiry assumes a somewhat 
speculative turn, as is obvious from the consideration that, despite 
the views put forward by Wallace 1 and others regarding the stability 
of the Continents, the inhabitable regions of the globe have cer- 
tainly undergone considerable modifications since the appearance 
of the Hominidae in their several geographical areas. Doubtless 
Wallace is right in rejecting Sclater's " Lemuria/ 1 Former 
as unnecessary to account for the range of the of^nd^and 
Lemurs. But he cannot reject the " Indo- African water. 
Continent," which replaces Lemuria in the Indian Ocean, and 
which is established on a solid foundation by the naturalists 
associated with the Indian Geological Survey*. Thus the hippo- 
potamus, now confined to Africa, is found in a fossil state both in 
Madagascar and in the Sivalik Hills (Himalayan foothills), while 
the plants of the Indian and South African coal measures are 
absolutely identical, and the remarkable Dicynodon and other 
allied forms of fossil reptiles are equally characteristic of both 
regions. Hence, although belonging mainly to secondary times, 
considerable sections of the Indo- African Continent, 
such as are still represented by Madagascar, the 
Chagos, Seychelles, Mascarenhas and other smaller Austral Conti- 

. , . nents. 

groups, must have persisted far into the tertiary 

1 The Comparative Antiquity of Continents, as indicated by the Distribution 
of Living and Extinct Animals, R. G. S. Journal, 1877, and elsewhere. 

2 See especially R. D. Oldham, The Evolution of Indian Geography, in 
Gcograph. Jour. March 1894. 


During this epoch Australia also was far larger than at present, 
not only comprising New Guinea, Tasmania, and perhaps New 
Caledonia in the Pacific, but also stretching westwards probably 
as far as the islets of St Paul and Amsterdam, that is, to within a 
relatively short distance of the Mascarenhas in the Indian Ocean. 
" The islands of St Paul and Amsterdam may indicate where an 
intervening land once formed a stepping-stone for the intermigration 
of the plants of Australia and South Africa." l In fact an Austral 
Continent dating from late secondary or early tertiary times, sur- 
viving in fragments down to the quaternary epoch, and extending 
from the Cape through Madagascar and Australia towards New 
Zealand, seems to be postulated by the huge cursores and other 
birds such as the sepyornis of Madagascar, the dodo of Mauritius, 
the Australian dromornis and the moa of New Zealand, surviving 
till quite recently in those regions. 

To the Indo- African Continent in the southern hemisphere 
corresponded a later (Miocene) Eurafrican Continent 
* n t ^ ie nortnern hemisphere, which occupied a con- 
siderable section of the present Mediterranean basin, 
as shown by the miocene formations, on the Mauritanian seaboard, 
in the islands and on the opposite side at intervals as far east as 
the Caucasus 2 . At that time the Sahara also formed, not a marine 
bed, as is generally supposed, but an elevated region at a much 
greater altitude above sea level than at present 3 . Thus in the 
miocene epoch there was continuous land almost everywhere 
between Europe and Africa, and the connection still continued at 
several points throughout pliocene times 4 , when gradual subsidence 
transformed the miocene plains first into three separate basins, 
and then into a vast inland sea extending from the Caucasus to 
the Atlantic. But geologically this marine inlet, on the shores of 

1 A. R. Wallace, Australasia (Stanford Series, new issue), p. 99. 
3 See F. W. Rudler and G. G. Chisholm, Europe (Stanford Series), 
Chap. I. passim. 

3 See A. H. Keane, Africa (Stanford Series, new issue), Vol. I. ch. Hi. 

4 At Gibraltar, where the present strait is of relatively recent formation ; 
between Tunis and Sicily, still connected by a shallow submarine ridge ; and 
between Libya and Greece united in tertiary times by a vast plain, the haunt 
of the lion and rhinoceros (Reclus, English ed. I. p. 36). 


which the tribes of men settled down " like frogs around a swamp " 
(Plato), is but a western extension of the still larger central Asian 
depression, which towards the close of the miocene age extended 
from Turkestan to Sicily, and the subsidence of which was syn- 
chronous with that of the Mediterranean, and with the final 
upheaval of the orographic system which stretched from the 
Pyrenees, through the Alps and Caucasus to the Himalayas. 

Thus when the pliocene precursor, wherever evolved, began to 
spread abroad, he was free to move in all directions over the 
eastern hemisphere. Like the anthropoid allied forms, he could 
have wandered, say, from the Indo-African Continent, either east- 
wards to India and to Malaysia, where are now the gibbon and 
orang, or westwards to Africa, where are now the chimpanzee and 
gorilla, and thence northwards to Europe whither he was preceded 
by the extinct miocene dryopithecus. From the Indo-African 
Continent the road was also, open through Australasia towards 
New Zealand, and from India to the shores of the flooded 
central Asian depression. Nor could climate anywhere present 
any difficulty, for this first dispersion took place 
during the long inter-glacial warm period, when a climate^ 1 *"* 1 
temperate flora ranged as far north as Spitzbergen, 
and when a rich arborescent vegetation afforded sufficient shelter 
from the fiery pliocene suns. 

From the Eastern Hemisphere the New World could at that 
time be easily reached either from Europe or from Asia. Without 
conjuring back Plato's vanished " Atlantis/' recent surveys have 
revealed the presence of a submarine bank, which stretches from 
Scotland through the Faroes and Iceland to Greenland, and 
which is nowhere more than 300 or 400 fathoms deep. Although 
partly of igneous origin, the corresponding strata on . 

both sides of the North Atlantic, together with accessible 
striking resemblances between the respective faunas rom urope * 
and floras, show that this ridge represents a vanished Continent 
of great age, which would appear to have still formed dry land in 
late tertiary times. Miocene limestone formations occur even in 
the island of St Mary, one of the Azores, midway between the Old 
and New Worlds, while Terceira, another member of the same 
group, is strewn with boulders both of crystalline and sedimentary 


origin, on the provenance of which geologists have however not 
yet made up their minds, whether transported by floating ice 
during the glacial period from the American mainland, or torn by 
volcanic agency from a subsiding continent. 

On the Asiatic side the two continents converge at Bering 
Strait to within a distance of 60 miles between 
^And from Cap^ East and West, while the strait itself has a 
mean depth of little over 20 fathoms. In clear 
weather the American is visible from the Asiatic headland, and in 
mid-channel lie the Diomede islets, stepping-stones between the 
two hemispheres. Farther south the Aleutian chain, enclosing the 
shallow Bering Sea, extends from the Alaskan Peninsula west- 
wards to the " Near Islands," so named from their proximity to 
the Siberian coast. For a great part of the year the intervening 
spaces are spanned by frozen masses, so that even before the first 
kayak was launched, primitive man might have passed on solid 
ground to and fro between the eastern and western hemispheres. 

But these essential factors, by which the problem, one might 
say, solvitur ambulando, have been for the most part either 
neglected or misunderstood by those who have approached the 
question of man's early migrations. Thus de Quatrefages, re- 
moving Sclater's Lemuria, without substituting the 

Migrations e ' & 

of primitive Indo-African Continent of the Indian Survey, 
man * leaves a great ocean flowing between the African 

and the Oceanic sections of the Negro division. Then, to meet 
the difficulty, he locates this with all the other primary divisions 
somewhere round about the Central Asian plateau, as if these 
groups could become differentiated in the same physical environ- 
ment, although to be sure, the conditions of existence are assumed 
to be different. The environment itself is reached from the 
Arctic regions where the precursor was evolved at a 
^ me wnen Spitzbergen enjoyed a temperate climate 
like that of California at present, and we are assured 
that this hypothesis of a boreal origin agrees with all the facts in 
the early history of man, and alone enables us to coordinate them 1 . 
From the extreme north tertiary man was driven en masse by the 

1 Histoire GMrale des Races Humaine$> I. p. 133. 


advancing ice-sheet to the central plateau, which is therefore taken, 
not as the cradle of the species, but as " le centre de caracteYisa- 
tion des types ethniques fondamentaux de l^poque actuelle" 
(p. 137). Yet in flat contradiction to this assumption it is added 
that it is not to be supposed that the migration southwards, deter- 
mined by the increasing cold, took everywhere the same direction. 
On the contrary some of the emigrants wandered away into 
America as far south as Brazil and the pampas, while others passed 
through Syria into Africa, sending offshoots (" e'claboussures ") 
south to the Cape. Central Asia thus ceases to be the offidna 
gentium, where the present fundamental types were elaborated. 
To the Negro division, and especially to the Negrito sub-group, is 
given an enormous expansion, radiating through Irania and South 
Arabia westwards to Africa, and through India south-eastwards to 
Oceania, these movements being required by the necessity of 
avoiding the Indian Ocean impassable before the invention of 
navigation. The general theory is supported by linguistic argu- 
ments, which are based on a radical misconception of the evolu- 
tion of speech. Thus it is affirmed that " d'une langue agglutina- 
tive ne sort pas un dialecte monosyllabique " (p. 300), the fact 
being that, as seen in Chapter IX., all monosyllabic languages 
have been developed from agglutinating forms. Again, the Negro 
migration from India to Australia is stated to be proved by the 
affinity of the Australian and Dravidian languages, " aujourd'hui 
universellement admise" (p. 333). This is one of those reckless 
assumptions which have brought philology into disrepute with all 
anthropologists, but respecting which it must suffice here to state 
that no sound philologist has ever affiliated the Australian to the 
Dravidian linguistic family 1 . 

The " Geographical Distribution of Mankind " has also been 
discussed by Mr James Dallas in a learned and 
well written monograph 8 , which, however, is also Da 
vitiated by a disregard for the distribution of land 

1 "The numerous Australian idioms seem all related to each other, but 
have no affinity with any other linguistic. family" (A. Hovelacque, Science of 
Language, English ed. 1877, P 67). "The Dravidian tongues may safely be 
regarded as an independent group, related to no other linguistic family 1 * (ib. 

P- 79)- 

3 Anthrop. Journal, 1885, pp. 304 30. In this essay Mr Dallas proposes 


and water in tertiary times. While the Indo- African Continent is 
ignored, the Sahara is submerged and Africa thus separated by an 
impassable liquid barrier from Europe. " Thus Europe would be 
effectually separated from Africa except at one point the Darda- 
nelles " (sic) ; and thus also the migrations not only of man, but 
also of the large African fauna into Europe would be left unex- 
plained. It is not surprising that the attempted scheme of 
distribution is almost admittedly a failure, and that the writer 
confesses himself "at a loss for a starting-point" for his "Meso- 
chroic " (Mongol) division. Here also linguistic and ethnical 
questions are confused, and a disposition is shown " to revert to 
the old Atlantis theory," in order to account for a purely fanciful 
" affinity of the Basque and American languages," an affinity 
which we are assured " must at once occur to every ethnologist " 
(p. 329). Basque has no affinities, beyond that due to loan words, 
to any other group in the New or the Old World, unless indeed 
G. von der Gabelentz can be said to have established a remote 
kinship with the Berber of North Africa. 

It may here be remarked that, however useful it may often be 
in connection with the study of existing races, language is of little 
or no avail in the elucidation of the early history of man. It is no 
longer possible to say how far the different present forms of 
speech had established themselves in those remote times; and 
such profound changes must have taken place since then, that 
resemblances between languages spoken thirty or forty thousand 
years ago have in any case necessarily long been obliterated. 
Some the Semitic for instance are no doubt marvellously persis- 
tent; but none, unaided by a written literature, could possibly 
resist the ravages of phonetic decay and other disintegrating 
influences acting for ages on the rude dialects of primitive man. 
Hence no use is here made of arguments drawn from linguistic 
resemblances or disparities, except only for the relatively later 
movements in the Indo-Pacific regions and elsewhere. 

Lastly, reference may be made to Dr D. G. Brinton's paper in 
the Forum for December 1 804 on " The Beginning of 

andBrinton. ^, , , 4 , _ ,...,, & & 

Man and the Age of the Race, which by a process 

Leucochroi, Mesochroi and dLthochroi as substitutes for White, Yellow and 
Black respectively. 


of elimination places the original home somewhere or anywhere 
along the southern slopes of the mountain ranges stretching from 
the Cantabrian Alps to the eastern Himalayas, but by preference 
in the western section, where " up to the present time his earliest 
vestiges have been exhumed.... Speaking from present knowledge, 
we must say we know of man nowhere earlier than within the area 
of England, France and the Iberian peninsula/' But all the 
known facts seem to imply that here man is an intruder arriving 
in west Europe from the south across the Mediterranean isthmuses 
(Ch. xiv.) in company with the great African fauna. West Europe 
is far too limited an area, and has been too frequently subject to 
upheaval and subsidence, to be the primeval home of the higher 
and larger mammals. But of course anything might happen any- 
where, according to this anthropologist's new and somewhat 
startling theory of " evolution per saltum? which is proposed as 
an alternative between the doctrine of " specific creation " and 
that of the " missing link," which is again made the butt of some 
needless ridicule. By this "evolution with a jump" is meant 
"that process which produces 'sports' in plants and 'cranks' or 
men of genius in respectable families... So it may have been with 
the first of men, &c." 

But, apart from these eccentricities, it cannot be denied that, 
although the missing link must be postulated (see p. 57), the 
failure, after a long and diligent search, to discover it in those 
regions where its presence might be looked for, is sufficiently sur- 
prising to need explanation. One obvious explanation may be 
that all traces of remote fossil forms must in any case be extremely 
rare, as seen in the few fragmentary remains hitherto discovered, 
for instance, of dryopithecus, which nevertheless must have 
abounded in the miocene forests of India and the Mediterranean 
basin. Unless protected by the accidental shelter of glacial 
deposits, rocky fissures and cavernous recesses, the osseous 
remains of animals strewn 'on the surface of the ground, or left 
undevoured by the carnivora, must with years crumble and mingle 
with the soil. Nor is it to be supposed that the search is ex- 
hausted, especially when it is remembered that scarcely a gene- 
ration has passed since inquiry has been turned in this direction 
by the appearance of The Origin of Species. 


But few missing links of much more simian aspect than those 

of Java or Neanderthal will probably ever be brought 

centre of to light, the pliocene precursor having apparently 

evolution. . . . , . , , 1^1 

originated in a now submerged area where the 
transitional forms can no longer be recovered. This area must 
obviously be sought in those regions of the Indo- African and 
Theindo Af- Austral Continents, which survived into tertiary 
rican and AUS- times, and which were the common home of the 
trai regions. anthropoids and of the lemurs with both of which 
sub-orders the Hominidae show affinities. It will be admitted that, 
cteteris paribus, such a region is more likely to have been the 
cradle of mankind than any other, where the lemuroid and anthro- 
poid precursors occur either only sporadically, or not in association, 
or else not at all. Thus are excluded, the whole of the New 
World and most of the northern section of the eastern hemisphere, 
leaving as the only possible centre of evolution some part of the 
southern section of the eastern hemisphere, where the proportion 
of land to water was far greater in the secondary and early tertiary 
periods than at present. In fact dry land extended continuously 
from the Atlantic to the Pacific, affording a free range to the 
lemuroids, the anthropoids and the dark Hominidae, all of which 
are now divided into western and eastern (African and Oceanic) 
groups by the intervening waters of the Indian Ocean. The true 
lemurs abound now mainly in Madagascar; but more generalised 
forms exist both in that surviving section of Indo-Africa (the 
gigantic Megaladapis and the Aye-Ay.? or Chiromys Madagas- 
cariensis)) in Malaysia (the Flying lemur, or Galeopithccus volitans), 
and even in Ceylon (the Loris or Nycticebidct, popularly known as 
"Slow Lemurs," found also in the Eastern Archipelago). So with 
the higher apes, as already seen, and with the two great sections 
of the Negro division of the Hominidse. The inference seems 
irresistible, that all these allied forms had their common primeval 
home in and about the Indo-African and Austral Continents, of 
which considerable sections still survive. 

Other considerations point with equal force in the same 

Characters of direction. That the immediate precursor was a 

the pliocene tropical or sub-tropical furry animal of arboreal 

habits is generally allowed, and this description 


applies to all the allied forms, some of which have coats combined 
of wool and sleek hair 1 . Man has both uncombined, and it is 
easy to see that, according to the requirements of the environment, 
one or other might be dropped, without assuming any transition 
from wool to hair. In other respects the precursor is described 
by de Quatrefages as probably red-haired, yellow-skinned, and 
prognathous, the red being perhaps rather a russet brown, the 
yellow a yellowish brown. This writer also points out that some 
of the sub-groups in the Negro division are not black but yellow, 
while the Negro himself shows a tendency to revert to this colour, 
whereas a tendency to hark back to darker shades is never 
observed in the yellow division and rarely in the white, from all of 
which phenomena it is inferred that blackness is not an original 
but an acquired character in the Negro division 8 . 

These views are confirmed by other considerations, such as the 
fact noted by Darwin that "the children of the Australians im- 
mediately after birth are yellowish brown and become dark at a 
later age 3 ," which is true also of the African Negro whose soles 
and palms are always yellow. With regard to the black hair both 
of the Negro and of the Mongolo-American, it is specially note- 
worthy that in East Tibet it is of a pale brown in ersistent in 
infancy, changing in the tenth or twelfth year to a the Afro-Aus- 
bright or glossy black, though in some cases a dark ra reglons> 
chestnut hue is retained for life, while the iris is always either 
brown or "d'un jaune fonce 4 ." Similarly Giovanni Pelleschi tells 

1 The Aye- Aye, for instance, "is clothed with longish smooth hairs with 
an under coat of a woolly nature " (N. S. Dallas, The Animal Kingdom, p. 772), 
somewhat like the lanugo of the human foetus. 

2 " On est conduit a admettre comme probable que nos premiers ancetres 
avaient la chevelure tirant sur la teinte rouge plus on moms roussatre. Le 
pigment cutane, qui donne aux individus et aux races leur couleur caracteris- 
tique, examine au microscope, presente toujours queique chose de plus ou moins 
jaune.... En invoquant encore les faits que je viens de rappeler, il est permis de 
penser que cette teinte dominait chez 1'homme primitif " (Op. cit. p. 156). 

8 Descent of Man, 2nd ed., p. 557. And according to Brough Smyth they 
"are nearly of the same colour as European children when born, and all of 
them are generally light-red" (The Aborigines of Victoria, I. p. 6). 

4 Desgodins, quoted by V. de Saint-Martin, art. Tibet, p. 591. 


us that the children of the Mattacco and Toba aborigines of Gran 
Chaco, Argentina, "up to ten or twelve years have reddish hair, a 
curious fact recalling the theory of De Salles, according to which 
primitive man was red-haired 1 ," like the Orang-utan of Malaysia. 
Even amongst the true Negroes of the Welle basin, Central Africa, 
"red hair occurs both amongst the dark and light peoples," while 
some of the dark Zandehs (Niam-Niams) have " very light, almost 
yellow-leather skins." fl The hair of the Wochua dwarfs in the 
same basin is described by the same observer as " of a dark, rusty- 
brown hue," and many are stated to have " full beards and hairy 
breasts " (ib. in. p. 82). Other Negritoes both in the western and 
eastern sections of the Negro domain, present more pronounced 
simian features than any other living human groups. Such are 
the Akkas of Mangbattuland, the Batwa and others of the Congo 
forest zone, the Sakais of the Malay Peninsula, the extinct Kalangs 
of Java, and the also extinct Australian tribe of the Adelaide 
district, whose skull, as described by Dr W. Wyatt 3 , reproduces 
the enormous superciliary arches and some other traits of Pithec- 
anthropus erectus and of the Neanderthal race. Thus are found 
still persisting or till lately surviving in these regions, and no- 
where else, several groups, which approach nearest both to the 
higher simian and to the earliest known human types. Some- 
of these groups, notably the yellow Bushmen of South Africa, the 
Sakais, the Aetas of the Philippines, the Karons of North-West 
New Guinea, and the extinct Tasmanians, have always stood at a 
stage of culture scarcely, if at all, higher than that of eolithic man 
in West Europe. 

Thus all the conditions point to these Indo-African and 
Pliocene and Austral lands as the most probable centre of evo- 
pieistocene lution of the pliocene precursor, who may have 
migrations. eas jiy migrated thence in small family groups to 
every part of the eastern hemisphere northwards through India 
to Central Asia, eastwards and westwards to Australasia and 

1 Eight Months on the Gran Chaco of the Argentine Republic, 1886, 
P- 3'- 

2 Junker, Travels, II. p. -240. 

8 Some Account of the... Adelaide and Encounter Bay Aboriginal Tribe &c. 
Adelaide 1879. This tri .be died out about the year 1850. 


Central Africa, and from Africa to Europe. From the already 
described distribution of land and water at that primeval 
time, it is evident that all the continents were home< 
directly accessible by " overland routes " to the migratory groups, 
which in their new homes became independently specialised by 
the natural process of readjustment to the differ- 
ent environments. And thus arose in the new V e? r pment of " 
centres of evolution the several pleistocene groups, the primary 

. groups. 

whence are derived without any violent transitions 
the present primary divisions of the human family. Treating of 
the relative antiquity of these divisions de Quatrefages concludes 
that " The human races have appeared in the following order : 
The Yellow, or at least a section of them, would appear to be 
the elder branch of the present human family ; other Yellow men, 
the Blacks and the Allophylian Whites followed apparently very 
soon after them, and it would be difficult to say which came first ; 
then may have come the Semites and at last the Aryans." 1 This 
successive evolution of Blacks, Allophylian and other Whites from 
different sections of a Mongol prototype, involves transformations, 
which are both improbable in themselves and unwarranted by the 
known facts. It seems far more natural to assume an indepen- 
dent and simultaneous evolution of the several pleistocene groups 
from a generalised pliocene precursor in different surroundings, 
where the specialised forms were each determined by their special 
environment, and afterwards diversely modified by fresh migrations 
and interminglings. Thus the question of "relative antiquity" 
scarcely arises, for all the present divisions ascend directly and 
independently in parallel lines from so many pleistocene groups, 
themselves determined by the physical conditions of their re- 
spective centres of evolution. By this assumption a reconciliation 
is also to a certain extent effected between monogenist and 

1 Op. cit. p. 1 6 r . By ' ' Allophylian Whites " are here meant those Europeans 
of fair type, such as the Finns and Basques, who are not of Aryan speech. The 
term "allophylian," from Gr. oXXos and 0vXi;, was introduced by Prichard 
(Nat. Hist, of Man , 2nd ed. p. 185), as the collective name of all European 
and Asiatic peoples not belonging to the Aryan, Semitic or Hamitic races. 
But like "Turanian" and other vague terms liable to be abused by popular 
ethnographists, it is now little used in strictly scientific. ethnological writings. 


polygenist views. The Hominidse are not separately evolved, in 
an absolute sense, that is, from so many different anthropoid pre- 
cursors ; but the present primary divisions are separately evolved 
from so many different pleistocene precursors, themselves evolved 
through a single pliocene prototype from a single anthropoid 

Such would also seem to be the assumption of Flower and 
Lydekker, who, in discussing the primeval disper- 

Flower and s j on remark that the first Hominidae were probably 

Lydekker on ' L J 

the spread of all alike (our pliocene groups) ; but as they spread 
over thegiob* over the globe, they became modified by climate, 
food, the struggle for existence with themselves and 
with other animals, by selection acting on slight variations, and so 
forth, the differences showing themselves externally in the colour 
of the skin, in the colour and texture of the hair, form of head and 
face, proportions of limbs and stature. These anthropologists 
also point out that geographical position must have been a main 
factor in determining the formation and permanence of races. 
Groups isolated in islands or secluded uplands would in due 
course develop new types in the physical and moral orders. But 
on large open spaces, continental plains or plateaux, unobstructed 
by great ranges or other natural barriers, free intercourse would 
make for uniformity. Smaller or feebler groups would be 
absorbed or wiped out, conquerors and conquered disappearing 
or merging together. " Thus for untold ages the history of man 
has presented a shifting kaleidoscopic scene," a ceaseless "de- 
struction and reconstruction," a constant tendency towards 
differentiation and towards fresh combinations in a common 
uniformity, the two tendencies acting against and modifying each 
other in diverse ways. At the same time the history of the evolu- 
tion of the present divisions has been mainly obliterated, and the 
absence of paleontological evidence, that is, of physical facts drawn 
from the remote ages when the different races were being slowly 
fashioned, makes their reconstruction largely conjectural. In 
other words, the geological record is necessarily imperfect, and 
many chapters being absent, the gaps between transitional forms 
cannot all be bridged over. The starting-point itself in the 
inquiry is unknown, and may never be discovered, as it may lie 


buried in the bed of the Indian Ocean, or of some other marine 
or lacustrine basin 1 . 

The detailed study of the several primary divisions, to which 
the following chapters are devoted, will tend to confirm these 
views regarding the geographical centres of evolution and disper- 
sion of the Hominidse. 

1 Introduction to the Stud} of Mammals, pp. 742 43. 

K. 16 



Two divisions: African and Oceanic Negro Family Tree The Negritoes: 
Two divisions Early migrations The African Negritoes The Akkas 
and Batwa The Bushmen and Hottentots Past and present Hottentot- 
Bushman domains The Oceanic Negritoes The Black element in India 
The Oceanic Negrito groups : Andamanese ; Sakais of the Malay 
Peninsula; Aetas of the Philippines; Karons of New Guinea; Kalangs 
of Java The Negro divisions compared The African Negro unprogressive 
without miscegenation Testimony of H. H. Johnston, Manetta, Ruffin 
and Sir Spencer St John Historic evidence Low state of Negro culture 
Two main sub-divisions : Sudanese and Bantu The Sudanese Negroes 
Mixed Sudanese groups The Fulahs The Negroid Bantus The Zulu- 
Kafirs and Wa-Huma The Bantu linguistic family General intermingling 
of the Sudanese and Bantu populations Hence classification impossible 
except on a linguistic basis Tables of the Sudanese and Bantu groups 
The Oceanic Negro domain An area of great ethnical confusion Two 
main sub-divisions: Insular Negroes and Negroid Australians Nomen- 
clature : Melanesians ; Papuans The Papuan domain, past and present 
The Papuan type The linguistic problem Wide diffusion of Malayo- 
Polynesian speech not due to Malay or Polynesian Migrations Still less 
to Melanesian Migrations The true explanation; the Caucasic factor 
The Australian sub-division Not homogeneous Constituent elements of 
the Negroid Australians and of the Tasmanians Tasmanian culture 

IN our Family Tree the " Generalised Negro " appears to be 
first detached from the parent stem. But strictly speaking it was 
not detached at all. The Negro group is to be conceived rather 
as remaining in the primeval home, left behind, so to say, while 
the others passed on to their several centres of evolution. As 

seen in the last chapter, this primeval home is 
divteions: assumed to be the Indo- Austral region now flooded 

African and by th e Indian Ocean. But before, or simultaneously 

with, the subsidence of the land, its human inhabit- 
ants gradually withdrew westwards to Africa, northeastwards to India 


and Malaysia, eastwards to South Australia and Tasmania and later 
to .New Zealand. Thus from the remotest times were constituted 
by easy and natural migrations the various Negro groups in those 
regions on both sides of the Indian Ocean, where they have always 
dwelt, and where they are still found, generally in association with 
the allied anthropoid apes. Perhaps the strongest argument for the 
original unity of all these groups, now separated by a great marine 
basin, is afforded by the fact that the two main sections, the 
African and Oceanic, comprise two distinct types, the tall Negro 
and the dwarfish Negrito. As the Negrito appears to represent 
the primitive stock, from which the Negro diverged 
later, such a parallelism cannot be regarded as a 
mere coincidence. In the accompanying Family 
Tree of Homo ^Ethiopicus are shown the main ramifications of 
both sections and sub-sections of the Negro stock. 

Here the parent stem, after throwing off the two great African 
and Indo-Oceanic branches to the right and left (west and east), 
soon dies out, submerged, as it were, by the rising waters of the 
Indian Ocean. That the Negrito 1 branches, from TheNe ri 
which the Negro proper is seen to break away at an toes : two 
early date in both regions, stand nearest to the 
primitive human type, seems self-evident, if de Quatrefages' 
description of the precursor be accepted as approximately cor- 
rect. It would also appear that the western (African) branch has 
on the whole preserved more of the original characters than has 
the eastern (Indo-Oceanic). Both no doubt present in certain 
groups (Akka, Sakai) an equal degree of prognathisrn, as well as 
an equally simian expression, combined with normally brachy- 
cephalic crania. But the African alone shows the original 
yellowish complexion, the reddish-brown woolly head, the some- 
what hairy body and the extremely low stature, ranging from about 
3 ft. 4 in. to a little under 5 ft. 2 . Few of the Malaysians fall much 

1 Span, negrito and negrillo, diminutives of negro; both forms occur, but 
negrillo is little used in English writings. 

2 Some of the dwarfs of the Semliki river between Lakes Albert and Albert 
Edward are spoken of by Captain Lugard as "about 3 ft. high," and reaching 
"to the hip-bone of Suron Adam, the Sudanese sergeant, who was about 
6 ft. 3 in." (The Rise of our East African Empire, II. p. 178.) But these do 

1 6 2 




below 4 ft. 6 in., while some, such as the Andamanese, rather ex- 
ceed 5 ft. The colour also is described as deep brown or blackish, 
so that it is not always easy to distinguish between the true Negri- 
toes and the Negroes (Papuans, Melanesians) of Oceanica; whereas 
in Africa no doubt ever arises. Here it may be remembered that 
the term "gorilla" was in the first instance applied by the Cartha- 
ginian Admiral, Han no, not to the anthropoid so named by 
du Chaillu, but to certain hairy women seen by him and his 
companions on the west coast, probably the dwarfs still surviving 
in the Ogoway basin. The Akkas, Wochua, and others of the 
Welle basin have a still more venerable historic record. They were 
not only known by repute to Aristotle, Herodotus, and even the 
Homeric singers, but had already been introduced 
into Egypt during the First Empire. At least Dr ti( ^s riy migra " 
W. Pleyte has shown 1 that the Akkas described by 
Miani and Schweinfurth most probably represent the pygmies 
sculptured on the tombs of Ti and Ptahhotep at Sakkarah, referred 
to the time of Tatkara (Tankheres) of the 5th dynasty, that is, 
according to Mariette, 3366 B.C. These figures, which are in bas- 
relief, faithfully reproduce their racial characters, while a dwarf 
from Beni-Hassan, in Upper Egypt, is depicted in Rossellini's 
design with the feet turned inwards, exactly like Schweinfurth's 
Akkas 8 . Mariette 8 points out that the Egyptians were acquainted 
with the Welle lands whence they procured these dwarfs, who are 
referred to in a hieroglyphic inscription recording that "to him 
come the pygmies of Niam-Niam from the Southern Lands, to 
serve in his household 4 ." Pleyte also mentions the well-known 

not appear to have been full grown ; and the Batwa of the district north-west 
of Luluaburg (South Congo basin) measured by Dr Ludwig Wolf, averaged 
quite 4 feet 3 inches (Nature, March 24, 1887, p. 497). None of the four 
Akkas brought to Europe in 1874 and 1876 (Marno and Long) exceeded 
3 ft. 4 in. 

1 Chapitres Suppttmentaires dit Livre des Marts ; Traduction ct Commen- 
taire, Leyden, 1883. 

2 "Us ne surpassent pas un metre de hauteur; ils ont les pieds tournes 
au dedans, ce qui rend leur marche chancelante " (n. p. 159). 

3 Soctiti Khtdiviale de Gtographie, April, 1876. 

4 From Dilmichen's Geographische Inschriften, PL 31, quoted by Dr 
Pleyte. It should however be stated that the hieroglyph transcribed Nam, 


statue of the dwarf, Nemhotep, who had a tomb of his own dating 
from the same 5th dynasty, and who belonged to the same group 
as those of the Sakkarah monuments. 

From Egypt, or else from Mauritania, where dwarfish tribes are 
still spoken of 1 , some of these Negritoes may perhaps have found 
their way into Europe in neolithic if not earlier times. At the 
meeting of the British Association, Oxford, 1894, the Swiss 
Anthropologist, Dr J. Kollmann, read a paper on "Pygmies in 
Europe," in connection with some human remains recently ex- 
humed from the neolithic stratum of a prehistoric station near 
Schaff hausen. Side by side with skeletons of the normal size were 
found four or five averaging not more than 1*424 mm., say, 4 feet 
8 inches. Reference was made in the same paper to the small 
people about 5 feet high still surviving in Sicily and Sardinia, 
that is, on the high road between pleistocene Africa and Europe, 
who were regarded by Dr Kollmann, not as degenerate Europeans, 
but as representatives of a distinct variety of mankind, which 
occurs in several types dispersed over the globe, and which he 
believes to have been the precursors of the taller races of man- 
kind. Some support is lent to this view by the folklore of many 
northern peoples, and perhaps even by more substantial evidence, 
such as the remains of little people said to have been found in 
the Hebrides by Dean Monro in 1549 and by the traveller Martin 
in about 1703, and in an island of Hudson Bay in 1631 by Foxe, 
who tells us that "the longest corpses were not above four feet 
long 2 ." But the Negroid affinities of all these pygmies is doubtful. 
Although many of the Akkas and some other groups are 

described as somewhat disproportioned and top- 
Nritoes. Can heavy, with tottering gait, the African Negritoes 

appear to be, on the whole, well made, except 
perhaps for a too protuberant paunch, very active, daring hunters, 
and fairly intelligent. Certainly the description given by Oscar 

or Niam by Dumichen is read Nu and Nun by Birch and Brugsch, and the 
term "Niam," now applied by their neighbours to the cannibal Zandehs, can 
hardly have been a territorial designation over 5000 years ago, 

1 A. H. Keane, Africa, 1895, I. p. 86. 

2 See Prof. B. C. A. Windle's Introduction to the re-issue (1894) of 
Dr Edward Tyson's Essay Concerning the Pigmies of the Ancients, 1699. 


Lenz of the Abongo of the Okande district, who are akin to 

Du Chaillu's Obongo of Ashiraland, and whom he 

speaks of as "physically and mentally degenerate 1 /' 

is by no means applicable to the Negritoes in general. They are 

in no sense a degraded race fallen from a higher state, but 

obviously a small people arrested in their upward development 

probably by an adverse environment. From time immemorial 


(African Negrito Type.} 

their home has been the great forest zone of Central Africa, where 
the original yellowish brown complexion may have been preserved, 
and where a short stature would be an advantage to a race living 
entirely by the chase, and thus compelled to pass their lives 
flitting about amid the tangled coils of tropical woodlands. 

E. G. Donnenberg, the only European who claims to have 
actually seen the Mauritanian dwarfs, speaks of them as "about 
four feet high, robust and well-made, and certainly not Moors or 
Berbers whose growth had been stunted by rickets, as they 
differed altogether from the other inhabitants of Marocco in 

1 "Physisch und geistig degenerirt" (Skizzen aus W. Afrika^ Berlin, 1878, 
ch. vi.). It is noteworthy that these Abongo are stated to be "very dolicho- 
cephalous" and of a "somewhat light chocolate- brown colour," whereas the 
Negritoes are normally brachycephalous and yellowish. 


physical appearance/' Junker's Wochua, south of the Akkas, are 
stated to be "well-proportioned, though the oval-shaped head 
seemed somewhat too large for the size of the body. In the 
upper jaw the facial angle showed a high degree of prognathism, 
and in those of lighter complexion the crisp hair was of a dark, 
rusty brown hue.... Hands and feet are of elegant shape, the 
fingers long and narrow, with relatively large nails. I found no 
trace of steatopygia and some other features characteristic of the 
Hottentots. All things considered, the Wochua must be regarded 
as normal (healthy) members of a wide-spread race of remarkably 
short stature, but otherwise fairly well-proportioned and well- 
developed. Hence they cannot be described as a morbid, 
degenerate people, as appears to be conjectured by Professor 
Ratzel 1 ." 

A very full account is given by Dr Ludwig Wolf of the Batwa, 
who may be taken as typical members of the Negrito 

The Batwa. f .. J . f _ ' TT 

family south of Congo. Here tney occupy numerous 
village settlements in the Sankuru and other river valleys, such 
settlements being met especially in the forest glades of districts 
inhabited by the Bakubu Bantus. They display wonderful agility 
both in climbing palm-trees to extract the sap, and in setting traps 
for game. In the chase they bound through the tall herbage 
"like grasshoppers," attacking the elephant and even the buffalo 
with their tiny arrows and darts. They are well made with 
absolutely "no deformity," averaging about 4 ft. 3 in. in height, 
with yellow-brown complexion distinctly lighter than that of their 
Bantu neighbours, short woolly hair and no beard. Dr Wolf 
unhesitatingly connects them both with the northern Akkas and 
with the southern Bushmen*, all being the scattered fragments of 
a primeval dwarfish race, who are to be regarded as the true 
autochthones of equatorial Africa. 

But whatever be their ethnical relation to the equatorial Negri- 
toes, there can be little doubt that the Bushmen 8 
The Bushmen. const i tute t ^ e aboriginal element in the whole of 

1 Travels, in. pp. 84-5. 

a "Nicht zweifelhaft erscheine" (Im Innern Afrikas, pp. 258-61). 
8 This term, which of course has no ethnical value, has been adopted by the 
English from the Dutch Bosjesman. The scattered groups have no general 


South Africa at least as far north as the Zambesi. Here they have 
been gradually driven to their present domain, the Kalahari 
Desert north of the Orange River, and Great Bushman Land, 
south of that river, by the Bantu populations advancing south- 
wards from the interior of the Continent. In some of their 
physical characters, as well as in their speech, they resemble the 
Hottentots, of whom some ethnologists regard them as a degraded 
branch, while others look on the Hottentots as a mixed race, 
resulting from unions between the Bantus and the Bushmen. 
Either view would satisfy many of the actual conditions, though it 
seems probable that they have suffered degradation in their present 
environment, where they have been hunted down like wild beasts 
by Boers and Bechuanas alike, and where they find little to live 
upon except game, snakes, lizards, locusts, roots, bulbs and 
berries. At times they pass several days in search of food, on 
which, when found, they gorge themselves, five persons devouring 
a whole zebra in a couple of hours. Their weapons are the bow 
and poisoned arrow ; their dress the untanned skins of wild beasts 
when procurable; their dwellings either the cave or a kind of 
"nest," formed by bending round the foliage of the bosje (bush), 
whence their Dutch name. They are grouped in small bands 
without any hereditary or elected chiefs, and consequently with no 
social organization. Even the family tie has become extremely 
loose, unions being of the most transitory nature. 

Yet, debased as they are almost to the lowest level of culture 
compatible with existence, the Bushmen are remarkably intelligent, 
and possess a sense of art far higher than that of the surrounding 
populations, as shown by the rock paintings of men and animals 
true to life found in their caves, and recalling the analogous repre- 
sentations of the Dordogne troglodytes. These rock drawings 
and paintings "differ much in aim and character. A large portion 
are of a caricature class, rudely but very spiritedly drawn in black 

designation, but call themselves Kwai^ which answers to the Hottentot 
"Men," and which supplies the plural postfix kwa t as in Saan-kwa (San- 
kwa, Soan-kwa\ the name by which the Bushmen are known to their 
Hottentot neighbours. Cf. the Hindi I6g 9 "people," also used colloquially 
as a personal plural ending, as in Admi-I6g, Mard-I6g 9 &c. According to 
Hahn, the word San means native, hence San-kwa= Aborigines. 


paint. The class representing fights and hunts is a large and 
interesting one.... Many of the drawings are representative of 
figures and incidents among white people, also of other native 
tribes. Some even suggest actual portraiture. The ornamentation 
of the head-dresses, feathers, beads, tassels, c., seems to have 
claimed much care, and to have given the native artists great 
pleasure in delineation. The higher class of drawings will be seen 
to indicate correct appreciation of the actual appearance of 
objects; and perspective and foreshortening are found correctly 
rendered V The Bushmen have also a rich oral folk-lore litera- 
ture, consisting of legends, fables, and animal stories in which the 
animals are made to talk each with its proper click, not otherwise 
heard in ordinary Bushman speech. These clicks, inarticulate 
sounds unpronounceable by Europeans, are peculiar to the Bush- 
man and Hottentot languages, the former possessing six, the 
latter four, three of which have been borrowed by the Zulu-Xosas, 
who have been for ages in close contact with both races. 

The Kalahari Bushmen are .described as taller and altogether 
a finer race than those of Cape Colony. But reports vary ; 
nor is it always easy to sift the evidence, for the term 
"Bushman" is often applied in a very loose way to dispos- 
sessed Hottentots, half-castes, or broken tribes owning neither 
flocks nor herds. "The Bushmen in Bechuanaland in the present 
day are following their masters' lead in the ways of civilisation. 
They are employed as herds and waggon servants, and on our 
recent journey to Shoshong we found on entering Khama's country 
that the chief had entrusted a flock of goats to the Bushmen who 
were living at Mamabula. In the heart of the Kalahari the 
vassals have flocks of goats of their own, while they herd also the 
flocks of their masters 2 . 1 ' 

Although the affinities between the Bushmen and Hottentots, 
both in physical type and speech, seem to be fun- 
tots 1 * Hctten " damental, the former present some sharp contrasts, 
especially in their more animated expression, their 
more furtive glance and more agile movements. The Bushman 

1 Notes on a Collection of facsimile Bushman Drawings , by Mark Hutchinson, 
Journ. Anthrop. Inst. 1882, p. 464. 

fl Rev. J. Mackenzie, Blue Book, 1885, p. 63. 


in this respect may be described as mercurial, the Hottentot as 
leaden, and the distinction applies with equal force to their mental 
qualities. Hence although occupying a much lower position 
socially, the Kwai appear to be endowed with a greater share of 
natural intelligence, and H. H. Johnston, like other observers, 
was much struck by the " mental ability " of the race, so "strangely 
at variance with their low physical characters 1 ." 

All things considered, it seems safe to regard the Hottentot 2 
as an intermediate form between the Bushman and the Negroid 
Bantu, but much more closely connected with the former than 
with the latter. This is seen, not only in their common speech, 
but also in their common yellow or yellowish brown colour, their 
abnormally prominent cheek bones, giving a triangular shape to 
the face, and some other peculiar racial characters, of which the 
tablier and steatopygia of the women are the most remarkable. 
But for the fact of their eugenesis both with Bantus and Euro- 
peans these traits might almost be regarded as specific, both 
appearing earlier in life and in a more exaggerated form in the 
Bushman, that is, the assumed original stock. In other re- 
spects the Hottentots are tall compared with the Bushmen (5 ft. 
4 or 5 in. and 4 ft. 8 in. respectively), with disproportionately 
small hands and feet (like the Negritoes), feeble muscular develop- 
ment, very broad flat nose, slightly oblique and deep-sunk eyes 
set wide apart, pointed chin, large lobeless ears, large mouth with 
thick pouting lips, pronounced prognathism (64 to 70), highly 
dolichocephalic head with very low cranial capacity (1290, Broca) 8 , 
short black woolly hair. The famous "Hottentot Venus" ex- 
amined by Cuvier, was really a Bushman woman, and consequently 
presented all these characters in a marked degree. " She had a 
way of pouting her lips exactly like that we have observed in the 
orang-utan. Her movements had something abrupt and fantastic, 

1 Jour. Anthrop. Inst. 1883, p. 463. 

2 Hottentot appears to be an onomatopoeic term invented by the early 
Dutch settlers to imitate certain recurrent sounds in the native language. 
Like their Bushman kinsmen, the people call themselves JChoi, "Men," or 
more fully, Khoi-Khoin, "Men of Men," and in some districts Hou-Khoin, 
"True Men," men in a preeminent sense. 

3 Dr Hermann Welcker gives for ten Bushmen 1240, but for ten Hottentots 
1369, which is higher than for many Negroes (Archivfiir Atitkrop. xvi.). 


resembling those of the ape; her lips were monstrously large... 
I have never seen a human head more like an ape's than that of 
this woman 1 ." This Hottentot Venus had a rival in prehistoric 
times in the "V&ms de Brassempouy," whose ivory statuette 
with several others was lately found in the undisturbed Quater- 
nary deposits of the Grotte du Pape at the station of Brassempouy 
in the Chalosse district, Landes. M. Ed. Piette, one of the 
explorers of the cave, describes these exceedingly realistic works 
of art as exhibiting physical characters (pronounced steatopygia 
and other features) analogous to those of the ancient inhabitants 
of Piint (Somaliland?) and of the present Bushman race. Whence, 
he asks, came these palaeolithic cave-dwellers, who were also 
distinguished by great hairiness, thick lips, the upper overlapping 
the lower, receding chin like that of the Naulette skull, and a 
remarkable development of fatty growth and excrescences about 
the pelvic region. "In quaternary times branches of the stock to 
which they belonged must have covered the whole of Africa and 
a part of Europe. In the Pharaonic epoch they were probably 
already extinct in Europe; but the allied races, although driven 
back and in a decrepid state, still occupy vast spaces from 
Somaliland to the Cape. The Egyptians, who knew them, have 
left us the portrait of the women of Piint, noted for their 
gibbosites fesstires. At present the adipose races are everywhere 
dying out, despite the taste of the Negroes, and even of the 
Berbers for voluminous forms. The Somali and the Bushmen 
still persist, though their inferior qualities place them at the 
lowest rung of the social ladder 2 ." 

Whatever is to be thought of this prehistoric diffusion of the 

Bushman or allied peoples, the former presence of 
entHotteSStl the Hottentot-Bushman elements all over South 
Bushman do- Africa is proved by the geographical nomenclature 

of the regions now occupied by the intruding 
Bantus. Thus the names of most water-courses contain some 
dialectic form of the word ib (ob, eb, op, iep &c.), which in 

1 Cuvier, quoted by Topinard, Anthropology, Eng. ed., pp. 493-4. 

2 La Station de Brassempouy et Us Statuettes humaines de la period* glyptique, 
in LI Anthropologie^ March-April, 1895. 


Hottentot means "water/' or "river," as in Gar-ib, "Great 
Water " (the Orange River), Hyg-ap, Nos-ob, Mol-op(o\ and 
others. The Wak-Wak of Edrisi's map (1154), which has so 
greatly puzzled historical geographers, may even be the Bushman 
Kwa-Kwa (Kwai-Kwai\ showing the presence of these abori- 
gines on the east coast south of Sofala, whence "long before the 
Portuguese circumnavigation of Africa they were driven back by 
Kafir tribes 1 ." 

Owing to these encroachments, continued for centuries, the 
Hottentot domain had been confined to the south-west corner of 
the Continent at the arrival of the first Dutch settlers in the lyth 
century. Since then it has been further reduced and broken into 
fragments by the development of European colonisation, so that 
at present the race is mainly represented by about 20,000 Namas, 
who give their name to Great and Little Namaqualand 2 , and who 
can alone be regarded as full-blood Hottentots. All the other 
groups, Hill Damaras, Koranas of the Upper Orange basin, 
Griquas of Griqualand West and East, and Gonaquas about the 
Kafirland frontier, numbering collectively about 180,000, are 
either Hottentot-Dutch or Hottentot-Negro half-breeds mostly of 
Dutch speech. The Namas alone still speak Hottentot, which is 
specially remarkable as one of the few languages of non-Caucasic 
peoples possessing grammatical gender and relational suffixes 
scarcely to be distinguished from true inflections. It shows no 
affinity to any other tongue except Bushman, although Lepsius 
felt inclined to group it with Ancient Egyptian on the ground of 
its highly developed grammatical forms. 

1 Dr Liechtenstein, Reisen, I. p. 400. So also Adelung and Vater : " Fur 
gewisse Gegenden ist diess vollig erweislich, indem Berge und Flusse des 
Landes, wo jetzt die Koosa [Ama-Xosa] wohnen, in ihren hottentotischen 
Namen den sichern Beweis an sich tragen, dass sie einst ein bleibender Besitz 
der Hottentoten gewesen sind" (Berlin, ed. 1812, in. p. 290). 

2 The qua of Namaqua is the above explained plural ending kwa. 
Damaraland farther north, which takes its name from the Dama-Herero 
(Hottentot-Bantu) half-breeds, should be Damaqualand. The ra is really a 
feminine dual form, so that Damaraland means literally "the land of the two 
Dama women." When the first explorers reached that region they asked its 
name, to which the guide answered Damara in reference to two native women 
visible at the time in the distance. 


As already seen de Quatrefages assigns a vast domain to 
the eastern (Indo-Oceanic) Negritoes, whom he 
represents as having left traces of their presence 
" depuis la Nouvelle-Guine'e jusqu'au Golfe Persique 
et des archipels malais au Japon," besides forming the substratum 
of the Dravidian and other populations in India and along the 
southern slopes of the Himalayas 1 . But, apart from vague refer- 
ences to Asiatic "Ethiopians" in Persia by Ctesias, Pliny, and 
other ancient writers, and to a dark element in Indo-China by the 
Chinese records, there is no proof at all of Negrito populations 
anywhere on the Asiatic mainland, except in the Malay Peninsula 
and possibly in India, precisely the very regions where they 
might be looked for. During his eastward migrations from the 
subsiding In do- African Continent, primitive man 
clement in would necessarily reach both of these regions, India 
India. directly, the Malay Peninsula through the Eastern 

Archipelago at that time forming part of the mainland, from which 
it is even now separated only by shallow waters scarcely fifty 
fathoms deep. Southern India itself is merely "the eastern half 
of a once more extensive land area," the gradual subsidence of 
which "took place during the last great period of earth-move- 
ments," which began towards the close of the miocene, and which 
"reached their maximum in the pliocene period 2 ," thus giving 
time for pliocene man to reach the Indian mainland* Hence the 
now generally admitted black substratum, forming the autochtho- 
nous element in that region, is no more than might be expected. 
Yet the real character of this element has given rise to much con- 
troversy, and owing to the absence of distinctly woolly hair, 
marked prognathism and brachycephaly amongst the low-caste 
aborigines of the Deccan, many ethnologists still deny the presence 
of true Negritoes in the peninsula. " Mop-heads " somewhat of 
the Papuan type are shown in a group of Veddahs of Ceylon 
photographed by M. de la Croix and reproduced by de Quatre- 
fages (n. p. 318). But it may be doubted whether any woolly hair, 
such as is common to all known African and Oceanic Negritoes, 

1 Races Humaines, 1 1. p. 351. 

2 R. D. Oldham, The Evolution of Indian Geography, Jour. Geo. Soc. 
March, 1894, pp. 176-7. 


has yet been seen in India proper. " The hair," writes Mr James 
Dallas, " is also black, but has never been stated with certainty to 
present the woolly character of the Negro ; but I would mention 
that to the best of my belief I have myself seen natives of India 
with unquestionably woolly hair. The reiteration of the contrary 
statement has, however, so unsettled my mind on the subject that 
I should now be loth to pronounce with certainty upon so simple 
a question 1 ." Fr. Miiller also tells us that "mention is every- 
where made of crisp (" gekrauselte ") often even of woolly hair 2 "; 
but the statement is too vague to decide anything. On the other 
hand E. Callamand describes the hair of the Mundas (aborigines of 
Baghalpiir) as "tantot lisses et raides, tantot frises," and this 
authority asserts that no woolly hair has yet been found in India, 
with a single doubtful exception ; he adds that the blacks of India 
are far removed from the brachycephalous Negritoes 3 . Still more 
conclusive is the evidence of F. Jagor and G. Koerbin, who made 
a careful study of 254 members of 54 low-caste and out-caste 
tribes of the Madras presidency, but failed to discover any woolly 
hair, all being either schlicht (straight), wellig (wavy) or at most 
kraus or gekrauselt (crisp or curly). The colour of the skin was 
mostly very dark, but never quite black, the darkest being "a 
somewhat shiny grey-black 4 ." Three only of the heads were brachy- 
cephalous, all the rest being either dolicho-, sub-dolicho- or even 
per-dolichocephalous, so that, all things considered, the dark 
clement in India would appear no longer to represent the original 
reddish-haired yellowish Negrito, but an intermediate form between 
that type and the Papuan, generally modified by later intruding 
Kolarian, Dravidian, and Aryan populations 5 . Referring to the 

1 Jotir. Anthrop* Ins. 1885, p. 308. 

2 Ethnographic, p. 139. 

8 Le Crdne des Noirs de V Inde> in Rev. cTAnthrop. Oct. 1878, pp. 607-625. 

4 Zeitschrift fur Ethnologic^ 1887, Part I. "Das ganz glanzende blauschwarz 
vermisse ich," the nearest being "em etwas glanzendes grauschwarz." 

8 It is noteworthy that M. Rousselet's portrait of a Jangali (properly Juang) 
approaches the Oceanic Papuan in the development of the nose and super- 
ciliary arches more closely than that of any other Continental dark type. 
These "Jungle people" who are said by Dr Caldwell to be the most primi- 
tive tribe in all India, live in the forest district a little north of Cuttack. They 
are represented by one skull in the Barnard Davis collection. 


worthless nature of the evidence relied on by de Quatrefages (in 
his work on the Pygmies) for the former wide-spread diffusion of 
the Negrito element, the late Professor V. Ball declares that he 
never met the slightest trace of this element amongst the 
numerous tribes visited by him "during many years' travelling in 
the hilly tracts of Western Bengal, the Central Provinces and the 
Northern Provinces of Madras. Individuals belonging to different 
tribes with curly, not really woolly, hair, are occasionally to be 
seen; but I venture to think that such occasional freaks are 
casual, wholly without significance, although they were regarded 
as evidence of a Negroid element in the population by the late 
Sir George Campbell 1 ." 

Hence we should no longer speak of Indo-Oceanic, but only 

The Oceanic ^ Oceanic, Negritoes, and even these differ in one 
Negrito material respect from their African congeners. The 

roups. original yellowish brown colour of the skin appears 

to have everywhere given place to various shades of dark brown 
and black, as amongst the surrounding Papuan populations. The 
Oceanic is even more fragmentary than the African domain, and 
the true Negrito element, formerly widespread throughout Ma- 
laysia, is now confined to the Andaman Islands, the Malay 
Peninsula, the Philippines and parts of New Guinea. A detailed 
description of the several groups would be foreign to the purpose 
of this broad classification, and a general survey with a view to 
establishing their racial unity must suffice. 

The Andamanese islanders, formerly spoken of as "Min- 
copies," present what Flower calls an infantile 

TheAnda- N e g ro type 8 , although in respect of stature they 


stand at the head of all Negrito peoples, averaging 
about 4ft. 10 in. Mr E. H. Man, who has made a special study 

1 Nature, May 23, 1895, p. 80. 

2 Osteology and affinities of the Natives of the Adamanese Islands^ in Jour. 
Anthrop. Inst. 1879, pp. 132-3. Here the Andamanese cranium is shown to be 
"as distinct as possible" from the Melanesian, and these islanders are spoken of 
" as representing an infantile, undeveloped or primitive form of the type from 
which the African Negroes on the one hand, and the Melanesians [Papuans] on 
the other... may have sprung," exactly in accordance with the views here 


of this race 1 , describes them as a homogeneous people, everywhere 
presenting the same uniform Negrito characters, short woolly 
black hair, very dark, almost black complexion, somewhat softened 
or undeveloped Negro features. They occupy a very low social 
state, living almost exclusively by hunting and fishing, in isolated 
groups of 50 to 80 persons, who wear scarcely any clothing and 
form both permanent and temporary camps of palm-leaf huts, 
varying in size and durability. They have names only for one 
and two, although able to count with the fingers up to ten, and 
otherwise show a considerable degree of intelligence as well as 
great affection for their women and children. Their social con- 
dition lends no support to the "cattle-herd" theory (p. 14), and 
the ferocious character formerly attributed to them is shown to be 
the reverse of the truth, and based on misunderstandings between 
them and strangers visiting the islands sometimes to kidnap the 
natives and sell them as slaves in the Malay markets. Since the 
British occupation none of the few recorded cases of hybridism 
have survived more than seven or eight years; the full-blood 
aborigines appear to be also dying out, numbering at present 
(1891) less than 4,000. The language, of which there are two 
distinct branches, is entirely unlike any other known form of speech, 
although in its morphology presenting certain analogies both to the 
Dravidian of India and to the Australian family. 

Geologically the archipelago is connected with the opposite 
mainland, so that migrations were formerly possible 
to the Malay Peninsula, where several small groups to 

of Negrito aborigines still survive. The Sakais, Malay Penin- 
Samangs, Jakuns, or Orang Bemia (" Men of the 
Soil ") as they are variously called by their Malay neighbours, are 
indeed more numerous than was formerly supposed, and, according 
to the Penang Administration Report for 1890, there may be over 
5,000 in the Ulu Pahang district alone. Here they form two 
distinct tribes, calling themselves Sen-oi and Tem-be, living for 
the most part in small groups of from two to three families, with 
little social organization. They speak a stock language, of which 
till recently little was known beyond the fact that it possesses 

1 In a series of papers contributed to the Jour. Anthrop. Inst, 1882-83. 
K. I? 

2 5 8 




(Oceanic Negrito Type.) 


(Oceanic Negrito Type.) 


names only for the first three or four numerals. But Mr Hugh 
Clifford has now made a study of the Sen-oi dialect, of which he 
publishes a glossary and grammar, with phonetic rules, showing 
no connection with any other known language 1 . There is almost 
everywhere much mixture with the surrounding Malay populations, 
resulting in many transitional forms. But the full-blood aborigines, 
as studied by Miklukho-Maclay, present the true Negrito type, 
even in an exaggerated form, with black woolly hair, dispropor- 
tionately large round head, and extreme prognathism. " This 
people undoubtedly belongs to the Melanesian stock 2 ." Special 
features are a very crisp black beard, a " third eyelid " or inner 
fold as in the Mongolic group, and the position of the three outer 
toes, which are turned obliquely towards the two inner, as in so 
many apes. This observer tells us that the Malays distinguish 
two groups, the Orang-Sakai-Liar, who are quite wild, keeping 
entirely aloof in the recesses of the forests, and the Orang-Sakai- 
Dind, who associate freely with the settled communities. One of 
Maclay's three photographs is described by Giglioli as presenting 
" a highly remarkable exaggeration of the'bestial characters, exceed- 
ing even the Kalang of Java in its prognathism... a real chimpanzee 
profile and I believe the highest degree of prognathism possible in 
a human being 8 ." 

Like those of the Malay Peninsula, the Negritoes of the 
Philippines, collectively known as Aetas 4 , are shown Thc Ne rf _ 
by Dr Blumentritt to be far more numerous than toes of the 
is commonly supposed. It also appears from Mon- lppm 

tano's recent explorations in Mindanao that they are very numerous 

1 Jour. Straits Branch R. As. Sac. No. 24, 1892. 

2 Ethnological Excursion in Johor. By "Melanesian" is here to be under- 
stood "Negrito," the Russian traveller habitually using the former term in a 
general way for all the dark Oceanic populations. 

8 "L* ultimo limite al quale possa giungere il prognatismo in un essere 
umano" (Nuove notizie sui Popoli Negroidi delP Asia e specialmcnte sui Negriti, 
Florence, 1879, P 7)- 

4 This term, which occurs in a great variety of forms Aeta, Aita> Atta, 
Ate, Eta, Ita, &c. has in the Tagala language the meaning of "black," being 
cognate to the Malay ^o*** (he*tam). Like the corresponding Mamdnua 
("Aborigines"), it is applied both to the full-blood and to the half-caste 

17 2 


in that large island as well as in some other districts, where their 
presence had not previously been suspected. But they are not 
always easily distinguished from the surrounding populations, 
many having adopted the dress and usages of the Malay intruders. 
Like the Sakais, many of the Aetas have formed close unions 
with these Malays, giving rise to various shades of transition between 
the two races, as shown in Dr A. B. Meyer's Album von Philip- 
pinen-Typen, Dresden, 1885. Many of the photographs in this 


(Oceanic Negrito Type.} 

collection are those of full-blood Aetas from Luzon and other 
parts of the Archipelago, showing the woolly hair, crushed nose, 
broad at base, deeply depressed at root, thickish and everted 
under-lip, sunken eyes set wide apart, long arms, slender extremi- 
ties, and wild look of the true Negrito. Some, especially of the 
children, have a distinct Negro expression, heightened by the low 
bulging frontal bone, so that they might well be taken for natives 
of Central Africa. In several a transition may be suspected 
between the Negrito proper and the Papuan, as might be expected 
from the position of the Archipelago on the confines of the 
respective domains. The same inference may be drawn from the 
physical appearance of the Karons, a group of Negritoes visited 

in 1879 b y M - Raffray, in the Arfak Hills, North- 
Guinea. W West New Guinea 1 . All alike are extremely rude, 

dwelling in wretched hovels of foliage and branches 

1 Tour du Monde^ xxxvii. 


and in some districts with no habitations, wearing no clothes 
beyond a few strips of bark dangling from a string round the 
loins, and (the Karons) addicted to cannibalism. " In the pure 
Negrito the height is said to average 4ft. xoin., but Semper's 
estimate is two or three inches less. The skull is brachycephalic, 
the chest small, the legs without calves, and the feet turned 
inwards. Their prognathous and deeply-lined faces give them an 
ape-like appearance. The nose is broad and flat, and the nostrils 


(Oceanic Negrito Type.) 

dilated, and the slender build and small size of the body cause 
the head to appear disproportionately large... Their intelligence is 
of a very low type, and according to Montano they are unable to 
count above five... They are monogamists without exception... 
Mr J. Barnard Davis, from the examination of three fine crania, 
considers the Negrito to be distinct from any other race 1 ." 

The Negritoes have left no traces of their presence in Formosa, 
if they ever reached that island, or in any other part of Oceanica 2 

1 Dr F. H, H. Guilleniard, Australasia^ vol. II. (Stanford Series, new 
issue) 1894, pp. 47-9. 

2 It should, however, be stated that during his scientific mission (1890-93) 
to Malaysia and Polynesia, Dr H. ten Kate collected what he considered 


except Java, home of the recently extinct Kalangs, in some 

respects the most ape-like of human beings. This 

T s h of Tavl" ma y ^ e inferred from the startlingly simian expression 

of Ardi, almost the last of his race, who lately died 

at Buitenzorg near Batavia, and of whom photographs have been 

preserved. Such a juxtaposition will cause no surprise, when it is 

remembered that Java must have been one of the first regions 


(Oceanic Negrito Tyfe.) 

reached by primitive man and his miocene precursor during their 
eastward migrations from the subsiding Indo-African Continent. 
Dr A. B. Meyer, who devotes a monograph to the subject 1 , speaks 
of a few of the Kalang tribe as still surviving, and Van Musschen- 

strong evidence of the former presence of Negritoes in Timor and the 
neighbouring islets of Samu, Roti and Savu, and especially in the Hokor 
district, Floras. From the appearance of the natives he infers that Timor was 
originally occupied by Negritoes, who were afterwards reduced and absorbed 
or exterminated by later Papuan intruders. He thinks with Crawfurd that 
here have been developed transitional types, the Negrito element prevailing in 
the west, the Indonesian in the centre of the island, though it is not quite 
clear what meaning this observer attaches to the term "Indonesian" (Tijd- 
schriftvan het Kon. Nederl. Aardrijkskundig Genootschap, Leyden, 1894). 
1 Die Kalangs attfjava, re-issued from the Leopoldina, August 1877. 


broek, to whom we are indebted for Ardi's photographs, informed 
Prof. Veth of Leyden, that " he has met with the same type in 
other parts of Java, though not so pronounced, and that it could 
always be traced to a Kalang origin 1 ." That they were the 
aborigines of Java gradually exterminated by the intruding Malays 
is not disputed, while Van Musschenbroek regarded them as akin 
to the other Negritoes of Malaysia. There could be no doubt on 
this point, but for the fact that when the photograph was taken 
Ardi's head was shaven, and since then Prof. O. Beccari, who 
saw him in 1878, found that the fresh growth was smooth, not 
woolly or frizzly, as had been expected 2 . What, then, is to be said 
of this Simian group, which is certainly not Malay, and presumably 
not Negrito ? It has been shown that the precursor was most 
probably furry, with a woolly under and a sleek outer coat, and it 
is conceivable that in a peculiar environment like that of Java it 
might have been advantageous to shed the wool .and retain the 
sleek hair, together with all the other physical characters of the 
primitive Negrito. Analogous processes are common enough 
especially amongst the ovidse, the European sheep changing its 
wool to hair in tropical lands, while in Sierra Leone all acquire 
black heads in a single generation 8 . No doubt the character of 
the hair, fixed by long ages, is now extremely persistent in the 
human varieties ; but it may have been less stable at an earlier 
period of their evolution. In any case it is readily modified by 
miscegenation, which might also be suspected amongst the mori- 
bund Kalangs now dispersed as menials and artisans amid the 
Malay populations. Only in that case the doctrine of correlation 
of parts would lead us to expect corresponding modifications in 
the other characters. 

Passing from the Negrito to the Negro proper, the most 
important point is the now established physical 
identity of the African and Oceanic branches. The Di %n 8 e * ro 
evidence bearing on this question has been sum- 
med up in a masterly manner by de Quatrefages, from whose com- 

1 Letter to A. H. Keane, Oct. 16, 1880. 

8 "I suoi capelli, cresciuti da quando fu fotografato, sono lisci" (E. H. 
Giglioli, he. cit. t p. 7). 

8 Winwood Reade, The African Sketch-book. 


parative craniological tables 1 are taken the subjoined broad 

results : 

Cranial Cephalic Facial Nasal 

Capacity Index Index Index 

S.W.Sudanese 130000. 6978 71*09 54' 

S.E.Sudanese 1355 71-66 71-09 54-16 

Mandingans 1460 72-82 68-18 54' 

Serrers 1490 6979 72-51 54-54 

Krumen 1445 72-28 69-16 51*92 

N. W. New Guinea 1305 71-11 71*42 55'io 

S. E. New Guinea 1385 71*89 69-92 53*56 

New Hebrides 1485 68*42 69-69 54*16 

Loyalty Is. 1460 69-84 68-38 51*92 

New Caledonia 1445 69-66 67-40 52*47 

Africans (mean) 1424-2 71*23 70-04 54*49 

Papuans 1412-5 70-38 68-87 53-03 

When to these anatomical resemblances are added such out- 
ward characters as a normally dark complexion, hair uniformly 
black and either frizzly or woolly in texture, the parallelism seems 
complete. Yet there are differences, such as the shorter stature, 
larger nose often arched and with downward tip, and generally 
milder expression of the Papuans, by which they may nearly 
always be distinguished at a glance from the African blacks. 

But the independent and simultaneous evolution of two types 
so nearly alike on either side of the Indian Ocean remains a 
remarkable phenomenon, which seems more than a mere coinci- 
dence, especially when the similarly independent or apparently 
independent evolution of two Negrito sub-types in the same 
regions is borne in mind. The explanation seems to be that both 
were already partly developed in the common centre of evolution, 
and after the dispersion east and west continued their evolution 
in the direction already taken. Then the observable differences 
would readily be accounted for by the influences of the different 
environments, both tropical, but one mainly continental, the other 
mainly Oceanic. 

These .differences are even more marked in the mental than in 

1 Races Humaines t II. pp. 319-20. 


the physical order. In some respects there is perhaps not much 
to choose between the two. Cannibalism was at no very remote 
period universal in both areas, although probably of a milder 
character in the east than in the west, where even since the 
" Partition " scenes of incredible brutality and atrocity have been 
witnessed in the Congo basin 1 . But the Papuan stands intellect- 
ually at a somewhat higher level than the African. He is less of 
an " overgrown child," more capable of social progress, less grossly 
superstitious, and possesses a much higher sense of Art, as seen 
by the splendid ethnographic collections recently made in the 
western parts of New Guinea by the agents of the Dutch Govern- 
ment 8 . 

Reference has already been made (p. 44) to the apparent 
incapacity of the full-blood African Negro to make The African 
any permanent advance beyond his present normal Negro unpro- 
condition without extraneous aid. In fact without 
miscegenation he seems to have no future, a truth which but for 
false sentiment and theological prejudice would have long since 
been universally recognised. Commissioner H. H. Johnston, 
than whom no better authority could be appealed to, fully agrees 
with the Negro writer who holds that "the pure and unadulterated 
Negro cannot as a rule advance with any certainty of stability 
above his present level of culture; that he requires the admixture 
of a superior type of man." But the white and Testimon 
black races "are too widely separated in type to ofH. H.John- 
produce a satisfactory hybrid." Hence he thinks S n> 
that "the admixture of yellow that the Negro requires should 
come from India, and that Eastern Africa and British Central 
Africa should become the America of the Hindu. The mixture 
of the two races would give the Indian the physical development 

1 The French explorer M. Fondese speaks of paddocks where "human 
cattle " were kept and fattened for the market, like stall-fed oxen. These were 
to be seen in almost every village in the Ubangi valley, and so resigned were 
the victims to their fate, that they actually refused the chance of freedom offered 
them by M. Fondese. 

* F. S. A. De Clercq and J. D. E. Schmeltz, Ethnograph ische Bcsehrijving 
van de West" en Noordkust van Nederlandsch Nieuw-Guinea, 1893. To these 
have now (1895) been added the collections made especially by Prof. A. C. 
Hadden in British New Guinea. 


which he lacks, and he in his turn would transmit to his half 
Negro offspring the industry, ambition, and aspiration towards a 
civilised life which the Negro so markedly lacks 1 ." 

In reply to those who attribute the backward state of the 
African Negro to baneful European and Mohammadan influences, 
it may be pointed out, first that Islam has on the whole been far 
more beneficial than injurious, as shown by the superior condition 
of those Sudanese populations, such as the Mandingans, Kansas 
and Sonrhai, who have been long in association with the Arab 
and Berber intruders; second, that the social status of the Negro 
masses is antecedent to all contact with European or any other 
foreign peoples. As already explained, their inherent mental 
inferiority, almost more marked than their physical characters, 
depends on physiological causes by which the intellectual faculties 
seem to be arrested before attaining their normal development. 
Even in the Southern United States under the plan- 

of Manetta, r 

tation system Filippo Manetta noticed that "the 
Negro children were sharp, intelligent, and full of vivacity, but on 
approaching the adult period a gradual change set in. The 
intellect seemed to become clouded, animation giving place to a 
sort of lethargy, briskness yielding to indolence. We must 
necessarily infer that the development of the Negro and White 
proceeds on different lines. While with the latter the volume of 
the brain grows with the expansion of the brain-pan, in the 
former the growth of the brain is on the contrary arrested by the 
premature closing of the cranial sutures and lateral pressure of the 
frontal bone 2 ." 

Has any real improvement taken place since the emancipation 
anywhere in the New World, where the conditions are more 
favourable than in the cradle of the race? After a lengthened 
experiment to raise the Virginian freedmen by education, involv- 
of Col Ruffin m % an ex P en diture f about ;i, 000,000, the late 
Col. Frank G. Ruffin finds the outcome to be that 
"so far from having been fitted by education for the discharge of 
civil or social duties, or from having been improved in conduct or" 

1 Report of the first three years' Administration of British Central Africa^ 
August 1894, p. 31. 

3 La jRazza Negra &c., Turin, 1864, p. 20. 


morals, they have absolutely deteriorated and have given no 
promise of amendment in any direction." This observer also 
notices "that negro children up to the age of puberty learn remark- 
ably well, at least by rote, but after that period of life has been 
reached they became incurably stupid and make no further 
progress V Hence " there has been no development of religious, 
intellectual, moral or industrial advancement in the Negro," who 
should be spoken of rather as non-moral than immoral, and who 
is here declared to be "a political idiot," an appreciation fully 
borne out by the results of a century of misrule amongst the freed- 
men and freemen of Hayti. Here the reversions to vaudoux and 
other pagan rites, to snake worship, cannibalism, and similar 
horrors are fully vouched for by Sir Spencer St John, 
who had official knowledge of these matters, and jj, r h ;* enccr 
who after a residence of over twenty years in " The 
Black Republic" was fain to confess that the greater his experience 
the less he "thought of the capacity of the Negro to hold an 
independent position. As long as he is influenced by contact 
with the white man, as in the southern portion of the United 
States, he gets on very well [?]. But place him free from ail such 
influence, as in Hayti, and he shows no signs of improvement; on 
the contrary he is gradually retrograding to the African tribal 
customs, and without exterior pressure will fall into the state of the 
inhabitants of the Congo. If this were only my own opinion, I 
should hesitate to express it so positively; but I have found no 
dissident voice amongst experienced residents since I first went to 
Hayti in January 1863 2 ." 

In Africa itself all social institutions are at the same low level, 
and throughout the historic period have made no 
perceptible advance except under the stimulus of 
foreign influences. Religion is a system of pure 
fetishism and ancestry-worship, associated with a universal belief 
in witchcraft and such sanguinary rites as those of the "customs" till 
recently practised in Dahomey and Ashanti. Slavery, 
where not checked by European governments, 
prevails everywhere both as a local institution and 

1 The Cost and Outcome of Negro Education in Virginia^ Richmond, 1889. 
3 Hay t^ or The Black Republic ', 1884, p. 131. 


a branch of the "export trade." The great bulk of the natives 
are still in the tribal state, while in the kingdoms founded in 
Upper Guinea, Ulunda, Buganda and elsewhere, the exercise of 
autocratic rule has nearly always been marked by the most wanton 
cruelties. The administration of justice is regulated, not so much 
by any sense of right or wrong, as by the caprice of the king, who 
is himself often in the power of the "witch doctor," Without 
external aid, no Negro people have ever reduced their language to 
written form, so that "literature" is purely oral, and limited to a 
few tribal legends, some folklore, proverbs, and songs of the 
simplest -kind. The arts are restricted mainly to coarse weaving, 
pottery, agriculture, wood carving, and the smelting and working 
of iron and copper, in which alone real skill and originality have 
been displayed. Architecture has no existence, nor are there any 
monumental ruins or stone structures in any part of Negroland 
except those of Sudan and Matabililand erected under Arab and 
Himyaritic influences. "No full-blood Negro has ever been 
distinguished as a man of science, a poet, or an artist; and the 
fundamental equality claimed for him by ignorant philanthropists 
is belied by the whole history of the race throughout the historic 
period 1 ." This is not the language of prejudice, of racial or 
religious bias, but the sober truth, frankly admitted by the Negro- 
philes themselves " behind the scenes." "In Massachusetts," writes 
Theodore Parker to Miss Hunt, "there are no laws now to keep 
the black man from any pursuit, any office that he will; but there 
has never been a rich Negro in New England..., none eminent 
in anything except the calling of a waiter 2 ." 

1 A. H. Keane, Encyclopedia Britannica, Art. Negro, pth ed. 

2 Letter, Nov. 10, 1857, quoted by J. R. Maxwell, almost the only "Negro 
of pure descent," as he calls himself, who has ever written a book (The Negro 
Question, 1892, p. 36). Dr Blyden, author of Christianity, Islam and the 
Negro Race, is also a Negro, or at least Negroid. No other instance has been 
recorded, although it is claimed for the Vei people of the West Coast that, like 
the Cherokees, they have invented an alphabet. The matter is involved in 
some mystification and needs further inquiry before the claim can be admitted. 
In any case it appears that the Vei are a branch of the Mandingans, who 
have been subject to Arab and Berber influences for nearly a thousand years 
(Capt. Binger, Du Niger au Golfe de Guinte, 1892, H. p. 213). 


On linguistic grounds the African blacks are conveniently 
grouped in two main sub-divisions, the northern 
SUDANESE, occupying a region of great linguistic 

confusion, and the southern BANTUS, amongst whom Sudanese and 
a remarkable uniformity of speech prevails every- 
where, except in the now contracted Bushman-Hottentot area. 
Sudan, which in its widest sense comprises the whole region 
stretching from the Sahara towards the equator, and from the 
Atlantic to the Red Sea, has with some reason been always 
regarded as the true home of the African Negroes, and in fact was 
so named from them by the medieval Arab writers '. This is the 
"Black Zone" in a pre-eminent sense, for here far 
more than south of the equator the Negro type is 
found in almost "ideal perfection," as amongst the 
Upper Guinea populations, the Serers of Senegambia, the Gallinas 
of Sierra Leone, the Sienufs within the Niger bend, the Mosgu of 
Lake Chad, the Fiir dominant in Ddr-FuY 2 , the Kordofan Nubas, 
the Dinkas and Shilluks of the Upper Nile, the A-Barambo, 
Zandehs and other of the Upper Welle Basin. During his excur- 
sion up the Nile Valley, the eminent French anthropologist, 
Dr E. T. Hamy, examined several specimens of Sudanese and 
Nilotic natives, presenting the usual Negro traits, such as great 
prognathism, high dolichocephaly and hypsistenocephaly, slender 
legs without calves, broad flat feet and larkspur heel (" talons forte- 
ment saillants en arriere"), and comparing these with observations 
made in other parts, he was satisfied as to "the indissoluble unity 
of the Western and Eastern Sudanese, a unity since then definitely 
confirmed to my mind by a large number of anatomical facts 3 ." 

But although Negro blood is almost everywhere dominant, 
Sudan, taken as a whole, is far from a homogeneous 
ethnical region. The greater part of the lands 
between the Nile and the coast are comprised within 

1 The full expression is Bildd es-S6ddn (bjJI &}), "Land of the 
Blacks," whence the terms Nigritia^ Negroland t figuring on all the old maps of 

8 Arab. Ddr, country, region &c., of frequent occurrence in East Sudan: 
Dar-Fur, Dar-Nuba, Dar-Fertit &c. 

8 Rev. cCAnthrop. 2 de se'rie, IV. 1 88 1, p. 225. 


the Hamito-Semitic domain; other Hamite and Semite (Berber 
and Arab) communities occur both in the east (Middle Nile, 
Kordofan, Ddr-Fdr, Waday, Lake Chad), and in the West (within 
the Niger bend and Senegal basin), while the ethnical " divides " 
are everywhere occupied rather by mixed Negroid than by full- 
blood Negro peoples. Such are, going eastwards, the Senegambian 
Toucouleurs, the Sonrhay of the Middle Niger, the Central 
Sudanese Hausas, Bornus and Baghirmi, the Mabas of Waday, 
the Base (Kunama), Barea, Shangalla and others of the Abyssinian 
slopes. The Base*, however, to judge from the figures reproduced 
by Mr F. L. James, represent an extremely low and even repulsive 
Negro type 1 . A difficulty is presented by the Fulahs, Moham- 
madan pastors, who were formerly dispersed in small 
communities throughout West and Central Sudan, 
but who, led by their warlike and fanatical chief, Othmdn Dan 
Fodio, rapidly overran nearly the whole region between Lake 
Chad and the Niger, and after overthrowing the native Hausa 
States (1800 1810), founded the present " empire 7 ' of Sokoto, 
with the vassal kingdoms of Gando, Nupe', and Adamawa. By 
some they are classed with the Negroes, by others with the 
Tuaregs (Saharan Berbers), while others again have brought them 
all the way from Malaysia. But this is not necessary, and when 
studied in their original homes the Futa-Toro and Futa-Jalon 
districts, Senegambia the Fulahs are found, despite their present 
Negro speech, to be of Hamitic type, possibly representing the 
Leuksethiopi ("White Ethiopians ") located by Pliny south of the 
Mauritanian Getulians. Grimal de Guirodon, who knew them 
well, describes the full-blood Fulahs as of reddish-brown or light 
chestnut colour, with crisp but not woolly hair, straight and even 
aquiline nose, regular features, and other characters separating 
them entirely from the Negro division*. Hence, despite *Fr. 
Miiller's " Nuba-Fulah Family/ they have no connection either in 
type or speech with the black Nubas of Kordofan. 

In the Bantu domain, which meets the Sudanese a little north 

1 The Wild Tribes of the Soudan, 1883; see especially the frontispiece, A 
^ Professional Beauty." 

2 Les Puls, 1887, passim. 


of the Cameroons on the west coast, and about the 

north end of Lake Albert Nyanza on the east side, B J n h t e us Negroid 

there are certainly some groups about the Lower 

Limpopo, Lake Tanganyika, the Ogoway and Lower Congo 

basins, which it is difficult to distinguish physically from the true 

Negroes. J3ut, speaking broadly, the Bantu populations show 

marked modifications of this type in their lighter colour, larger 

cranial capacity, smaller teeth and less pronounced prognathism. 

(Bantu Type.} 

They are also distinctly more intelligent, more civilised, and more 

capable of upward development than the full-blood Negro. The 

Zulu-Xosas (Zulu-Kafirs) of the extreme south-east, 

who stand out conspicuously in all these respects, K Jfi h r s ZulU " 

are taken as typical members of the division, and 

from their language has been adopted the term Bantu (properly 

Aba-ntu, "people 1 ") now used as the conventional name of all 

1 Aba is one of the numerous plural personal prefixes, each with its corre- 
sponding singular form, which are the cause of so much confusion in Bantu 
nomenclature. To aba, ab, ba answers a sing. ;;///, um, mu, so that sing. 
umu-ntu, um-ntu or mu-ntii, a man, a person; pi. aba-ntit, ab-ntu, ba-ntu. 


African races of Bantu speech. These are essentially mixed 
Negroid peoples, the dominant element being undoubtedly the 
Negro, as shown by the universal prevalence of black woolly hair 
and dark complexion, besides gross superstitions associated with 
witchcraft of a specially Negro character. With the black sub- 
stratum are intermingled Semitic (Arab) intruders on the east 
coast, and elsewhere most probably Hamites, chiefly Gallas, 
descending from the north-east. The so-called Wa- 
Huma*" Humas, dispersed amongst the equatorial lake popu- 
lations, with whom they are slowly amalgamating, 
are known to be Hamitic Gallas *. The founders of the Kitwara 
empire, now broken into fragments (Buganda, Bunyoro, Karagwe*) 
were also Gallas, as is evident from the fact that Galla was the 
mother-tongue of the late King Mutesa of Buganda, a lineal 
descendant of the Kitwara dynasty. A distant branch of the 
same race are the fierce nomads of Masailand, east of Victoria 
Nyanza, though probably modified by a strain of black blood, and 
the same process of segmentation and infiltration has obviously 
been going on for ages, leavening the seething masses throughout the 
southern half of the continent, and raising them to a somewhat 
higher level than that of the full-blood Sudanese aborigines. 

Hence in the Bantu domain every shade of transition is 

The Bantu presented between the extreme Negro and Hamitic 

linguistic types ; hence also the impossibility of determining 

family. ^ clearly marked Bantu physical type, so that this 

term has rather a linguistic than an ethnical value. It thus 

But in some of the groups mu is also plural, the chief dialectic variants being 
Ama, Aba, Ma, Mu, Ba, Wa, Ova, Va, Vua, Ua, U 9 A, O, Eshi, as in Ama* 
Zulu, Mu-Sarongo, Ma-Yomba, Wa-Sivahili, Ova- Her era > Vua~Twa> Ba~ 
Suto, Eshi' Kongo. Equally numerous and perplexing are the class pre- 
fixes indicating speech: Ki, Kishi, Di, Lti, So, Se &c., as in Ki-Swahili, 
Kishi-Kongo, Lu-Ganda, Se-Suto,=ihe Swahili, Kongo, Buganda and Basuio 
languages. It would be well if the Swahili Wa and Ki were universally 
adopted, as is the practice of some writers. 

1 Thus Stanley speaks of the Wa-Kerewe islanders, Victoria Nyanza, as 
*a mixture of the Ethiopic [Hamitic] and Negro type" (Through the Dark 
Continent, I. p. 251), and in Usongora he met certain Wa-Huma chiefs who 
" were as like in features to the finest of the Somali types and Wa-Galla as 
though they were of the same race " (In Darkest Africa, II. p. 317). 


corresponds to such names as Aryan, Mongolo-Tatar and Malayo- 
Polynesian, which similarly imply linguistic unity amid much 
physical diversity. As far as is known and the region has now 
been almost everywhere traversed by explorers all the innumer- 
able dialects current throughout the Bantu domain are more or less 
closely related in structure, phonetics and vocabulary, and have all 
certainly sprung from a common Bantu mother-tongue, differing 
fundamentally from all other known forms of speech. This stock 
language is distinguished by some remarkable grammatical features, 
of which the most characteristic is a certain alliterative harmony, 
somewhat analogous to the vocal harmony of the Finno-Tatar, 
and the nominal concordance of the Aryan system. The allitera- 
tion is caused by the repetition, in a slightly modified form, of the 
same prefixed element before all words of the sentence in gram- 
matical concord. Hence inflection in Bantu is mainly initial, not 
final, as in most other systems. All nouns are grouped in so many 
classes, according to their proper determining prefixes, of which 
there appear to have been at least sixteen in the organic Bantu 
language ; it follows that all adjectives and other words dependent 
on the noun are liable in principle to sixteen initial changes, 
according to the several classes of nouns with which they may 
occur. Thus : umu-ntu om-kulu, a great man, but in-kose en-kulu, 
a great chief, where kulu, great, becomes om-kulu, en-kulu,... in 
agreement with umu-ntu, in-kose... Compare Lat. domin-us bon-us, 
domin-a bon-a, &c. The germs of this concordance, which gives 
the clue to grammatical gender in the inflecting orders, are found 
in Masai, Galla, Tibu and some of the Nilotic tongues. Traces 
of alliteration depending on the same principle occur also in some 
of the idioms of the Welle basin and elsewhere in the border lands 
between the Sudanese and Bantu areas. But the principle is fully 
developed only in Bantu, which would thus appear to have origi- 
nated in the north, and to have spread thence with the prehistoric 
Hamitic {Galla) migrations throughout South Africa. How 
rapidly a Bantu language may be diffused by such migrations is 
seen in the case of the Makololos, a Basuto people who about 
1825 moved several hundred miles northwards to the Zambesi, 
where they reduced the dominant Barotse nation and founded a 
powerful state under their renowned chief, Sebituane. Then the 
K. iS 


Barotse suddenly rose (1864) against the intruders, exterminating 
them almost to a man, and restoring the old Barotse kingdom. 
But although the invaders have disappeared, their Sesuto language 
still survives as the current speech throughout the Upper Zambesi 
basin 1 . 

Throughout the historic period a great part of Negroland has 
been wasted by similar hostile movements, con- 
s pi cuous amongst which were the widespread 

the Sudanese expeditions of the terrible Jagas in the i yth century. 
populations. Scarcely less destructive were the kidnapping raids, 
^dating back to the old Egyptian Monarchy, revived 
by the Western nations to supply the hands needed to work the 
mines and plantations in the New World, and continued down to 
the present time by the Arabo-Nubian slave-hunters and their 
native allies. The result was an incessant dislocation, breaking 
up and re-formation of the tribal groups, and a universal inter- 
mingling of the most diverse elements, so that the utmost ethnical 
confusion now prevails throughout both the Sudanese and the 
Bantu domains. In fact hopeless chaos would seem to have been 
prevented mainly by the principle of convergence, which con- 
tinually tends towards uniformity of type in a given environment, 
thus to some extent counteracting the influences which tend in the 
opposite direction towards divergence. Hence the broad general 
resemblances already noticed in these regions, although even 
within comparatively narrow areas great diversity has often been 
observed by intelligent travellers. Thus Junker speaks of the 
" endless gradations of colour " on both slopes of the Nile-Congo 
waterparting, " ranging from the rarely-occurring deep black to a 
dark iron-grey, dark chocolate or roasted coffee-berry, light cigar, 
the yellow-brown of dressed leather, cafe-au-lait, and, in exceptional 
cases, the fair colour of the Malays." He adds that " red hair 
occurs both amongst dark and light peoples 2 ," as in the other 
primary divisions. 

How is it possible, after these long continued tribal inter- 
minglings, to speak of any scientific classification of the second- 

1 Livingstone, Travels \ Holub, Sieben Ja hre in Sud-Afrika, 1881. 

2 Travels, n. p. 240. 


ary divisions? Refuge is naturally taken in differences and 
resemblances between languages, which, as seen, do not inter- 
mingle, and which under certain conditions may 
have some value. Thus the Gold and Slave Coasts Hence Class- 

... . , , _, __ ification im- 

are occupied by a considerable number of Negro possible except 
tribes speaking three or four marked dialects of a Sa S f s . hnBUistic 
common stock language Tshi, Ga, Ewe, and Yoruba 
and also, as shown by Ellis 1 , presenting numerous points of 
resemblance in their physical characters, social usages, religion, 
traditions and progressive grades of culture. It seems reasonable 
in such cases to infer common genetic descent also. Analogous 
instances occur in other parts of Sudan, as amongst the Sonrhay, 
who may be traced by their speech from within the Niger bend 
eastwards to Asben, which district is known to have formed part 
of the powerful Sonrhay empire overthrown by Marocco in the 
1 6th century. So with the Fulahs, who can be followed by means 
of their language throughout all their wanderings from near the 
Atlantic seaboard right across the Black Zone to 

TN f -i- / i i i i i j- *. t Tables of the 

Dar-Fur, although no longer everywhere distinguish- Sudanese and 
able by their physical features from the surrounding Bantu g rou P s - 
Negro populations. Hence in the subjoined Tables of the 
Sudanese and Bantu peoples, the groupings have necessarily to a 
large extent a linguistic base. 


Wolofr between Lower Senegal and Gambia; chief branch 
Jolof; very black, but somewhat regular features, showing Hamitic 

Serer, Salum river and Cape Verde district; tallest of Negroes, 
many 6 ft. 6 in. ; herculean frames. 

Toucouleur (Tacuror), Kaarta district and Senegal river; a 
historical people formerly powerful in W. Sudan ; Negroid. 

1 The Tshi-Speaking Peoples of the Gold Coast, 1887; The Ewe-Speaking 
Peoples of the Slave-Coast, 1890; The Yoruba- Speaking Peoples of the Slave- 
Coast^ 1894. 

2 Assumed to be approximately full-blood Negroes where no indication is 
given to the contrary. 

1 8 2 




Mandinga?^ the chief nation in W. Sudan, with numerous 
branches between the Upper Niger and the Coast : Sarakole', 
Kassonke', Jallonke', Suzi, Susu, Vei, Solima &c. ; mostly Negroid. 


Khabunke, Balanta, Bagnum, Upper Casamanza and Cacheo 

Felup, Casamanza and Cacheo estuaries. 

Landuman, Nalu, Baga, Sape, Rio Nunez basin. 

Bullom, Mendi, Limba, Gallina, Timni, Sierra Leone. 

Pessi, Gola, Kondo, Basso, Kru, Webo, Liberia. 

Avekvoni) Agni, Oshiu, Ivory Coast. 

Tshi, Ga, Ewe, Yoruba, Gold and Slave Coasts. 

Sonrhay (Songhay) Middle Niger, and east to Asben. 

Hausa, the chief nation between the Niger and Bornu. 
Negroid; speech shows Hamitic influences. 

Bolo, Yako, Tangala, Kali, Mishi, Doma, Benue basin. 

Igarra, Ibo, Iju, Okrtka, Nempk, Niger delta and Oil Rivers. 

Efik, Qua, Andoni, from Bonny to Rio del Rey, where Bantu 
domain begins on the west coast. 

Borgu, Garma, Mossi, Tombo, Gurunga, Sienuf, within the 
Niger bend. 


Kanuri, Bornu, Negroid ; speech shows Tibu influences. 

Baghirmi, Lower Shari basin. 

Mosgu, between Lake Chad and Adamawa. 

Yedina, Kurt, Islands in L. Chad. 

Maba, Birkit, Massalit, Korunga^ Kabbaga &c. ; Waday, 
mostly Negroid. 

Fur, Kunjara, Tegele, Dar-Fur, Kordofan. 

Nuba, Kargo, Kiilfdn, Kolaji, Tumali, Kordofan. 

Nubians, Nile Valley between Meroe and Egypt 1 . 

Shuli ; Labore, Luri, Bari, Nuer, Dinka, Shilluk, Mundu, 
Abaka, Bongo, Mittu, Golo, Tonj and others, Upper Nile and its 
western affluents between Lake Albert Nyanza and the Sobat con- 

Kirim, Ishing,Janghey, Bonjak, Komar, Sobat basin. 

Kalaka, Mangbattu (Mombttttu), A-Bangba, A-Madi, A-Zandeh 
(Niam-Niam), Momfu, A-Kahle, A-Barambo, A-Babua, Embata, 
Mangballe, A-Banjia, Mabenge, Nsakkara, A-Ngaddtt, Welle basin 
from source to Mbomu confluence 2 . 

Bast (Kunama), Barea, Mareb basin, Upper Nubia. 

Shangalla, Gambil, western slopes of Abyssinia and Gallaland. 


Bayon, Ndob, Basa, Balun, Abo, Barombi, aborigines of the 
Cameroons 8 . 

1 For the intricate relations between the Negro Nubas of Kordofan and the 
Negroid Nubians of the Nile valley see A. H. Keane, Ethnology of Egyptian 
Sudan^ 1884, pp. 12 16. Here also the reasons are given for rejecting 
Fr. Mailer's "Nuba-Fula" Family. 

2 In this borderland between the Sudanese and Bantu areas there is a great 
intermingling of tribes. From what little is known of the languages (ten 
vocabularies collected by Junker) Leo Reinisch infers a distant connection 
with the Bantu form of speech. The aboriginal Negro element seems to be 
best represented by the A-Kahle of the Mbomu affluent, who "probably 
occupy their present domain from remote times," and who "are the only 
nation that has not suffered dismemberment". (Junker, in. p. 280.) 

8 This distinction, made by H. H. Johnston, between the aboriginal and 
intruding Bantus in the Cameroons territory, **is based, not on physical 
appearance, which is nowhere sufficiently marked for purposes of classification, 
but on linguistic grounds, the indigenous tribes speaking archaic Bantu idioms 



Barondo, Bafarami, Bakundu, Bamboko, Bakwiri, 
Duala, Bakoko, Banoko, Bapuko (Great Batanga), Bafindi, Ibea, 
intruders in the Cameroons from the east and south 1 . 

Bubi (Adeghaz), Fernando Po, Bantus in speech only*. 

Fan, intruders in the Gaboon and Ogoway basins 3 . 

Mpongwe, Gaboon estuary. 

Mbenga, Corisco Bay and islands. 


Galboa, Mnga, Okanda, Apinji, Ashango, Ishogo, Lower and 
Middle Ogoway basin. 

Oshebo, Aduma, Osaka, Mbamba, Upper Ogoway basin. 

degraded by long contact with their Negro neighbours, while all the later 
arrivals except the Ibeas speak comparatively pure Bantu tongues connected 
by imperceptible transitions along the seaboard with those of the Lower Congo " 
(A. H. Keang, Africa, 1895, II., ch. i). 

1 See previous note. 

2 "Les distinguent tres nettement de toutes les tribus cotieres 
par les traits, par la couleur jaunatre de la peau, par les cheveux, qui sont longs 
et frises, mais nullement laineux" (De Quatrefages, Races Humaincs, II. 
p. 404). 

3 A cannibal people who reached the west coast from the interior during 
the ipth century, and who are described by Burton, Oscar Lenz and other 


Bateke, Apfuru, Alima tributary of Lower Congo. 

Cabinda, Mayombe> Bakamba, Kuilu basin and thence to 
Congo estuary. 

Bangala, Mayakka, Pfl/&/<?&7, Kwango basin. 

BakutUy Bakuba, Bakete, Tushilange> Baluba^ Balolo, Eshi- 
Kongo, southern affluents Middle and Lower Congo. 

Ababambo, Abanja> Ubangi valley. 

Babanda, Babesse, Banalya, Aruwimi valley. 

Vuaregga, Yambarri, Manyuema, Vuarua, Basamba^ Congo 
basin above Stanley Falls. 

Kalunda, Vuabisa, Vuarunga, Vuafiba, Uvmza^ Vuahha, 
Lakes Moero, Bangweolo and Tanganyika. 

Abunda, Quissama^ Amboella, Angola, Benguela. 

Ovampo, Ova/iererOj Damaraland. 

Amaxosa, Amatembu, Amampondo, Amafingu, Amazulu, 
Matabili, Maviti, Cape Colony, Natal, Matabililand, Nyassaland. 

Bechuana,) Basuto, Makalaka^ Mashona, Banyai, Bechuana, 
Matabili and Mashona lands. 

Ganguella> Baviko, Barotse, Mambunda^ Kubango and Upper 
Zambesi basins. 

Batonga, Bashukulumbwe, Kafue and Middle Zambesi basins. 

Wankonde, Manganja> Wayao, Nyassaland. 

Magwangwara,) Makua, Maviha, Mozambique. 

Makonde, Wazaramo^ Wasagara, Wagogo, Vuazinza^ Wasam- 
bara, Wanyamwezi, Waswahili y between the east coast and 

Wateita, Wataveita, Wachaga, Kilimanjaro district. 

Wapokomo> Tana basin, conterminous with the Hamitic 
(Somal, Galla) area. 

observers as quite distinct from the surrounding Negroid populations, of light 
brown or yellowish colour, full beard, tall slim figure and very prominent 
frontal bone. Lenz (Skizzen, p. 35) describes the language as ** entirely 
different from that of the other Negro tribes," whereas Winwood Reade 
(Sketch-book, I. p. 108) says that " it is like Mpongwe (a pure Bantu idiom) 
cut in half; for instance njina (gorilla) in Mpongwe is nji in Fan." This 
word Fan itself, meaning "Man," is stated to be cognate with Bantu, and the 
plural is formed in the usual Bantu way: 2fo-/tt*=Men. It occurs in several 
forms, Pahuin (adopted by the French), Pamue, JFanwe, Mpangwe, &c. 


Wasoga, Waganda,) Wanyoro^ Victoria and Albert Nyanzas. 
Wakonjo, Wambuba, Wawamba, Walenga, Lake Albert 
Edward and Semliki basin; Ruwenzori 1 . 

In the Oceanic Negro division, where intercourse has always 
been facilitated by the prevailing trade winds and 

ro omain marme currents, racial interminglings have taken 
an area of place even to a greater extent than on the African 

ethnical con- . , , r^. r ,, ,, , 

fusion. mainland. The confusion of types is all the more 

perplexing in that this watery domain has from the 

remotest times been easily accessible from the southern shores of 

Asia, with which it still formed continuous land probably so 


recently as the pleistocene age, when that Continent would appear 
to have been already occupied both by Mongolic and Caucasic 
peoples. From their prehistoric migrations to Malaysia, Australia, 

1 Here the Bantu and Sudanese domains appear to overlap, and Dr Stuhl- 
mann, who explored this region in company with Emm Pasha in 1891, speaks 
of the Wakonjo and some other local tribes rather as full-blood Negroes than 
negroid Bantus (Petermanvfs Mitteilungen^ June, 1892). 


and Polynesia have arisen some difficult ethnical problems, which 
will be discussed farther on. Here it should be noted that all 
these regions, now occupied by so many different races, were 
the primitive home of the Oceanic or eastern branch of the 
Negro division; consequently that the black is everywhere to 
be regarded as the aboriginal element, the others as later in- 

Besides the already described Negritoes, this black element 
comprises two broad sub-divisions, presenting such 
marked physical and mental differences that no sub-divisions: 
systematist has ventured to group them under a ground? 6 " 
common designation in the same category. These Negroid AUS- 
are the insular blacks, true Negroes, whose domain 
originally comprised the whole of Oceanica, taken in its broadest 
sense, and the Continental Negroid blacks, comprising all the 
aborigines of Australia with the extinct Tasmanians. The distinc- 
tion is thus somewhat analogous to that which was seen to obtain 
between the Sudanese Negroes and the Negroid Bantus of the 
African division. The parallelism is even closer than might 
appear from this statement, as will presently be seen. 

No quite satisfactory general name has yet been proposed for 
the insular blacks, who are commonly referred to either as Mela- 
nesians or Papuans. But the use of Melanesians in this general 
sense gives rise to much confusion, as the term has 

. , , . . . . .. . , Melanesians. 

a long-established special meaning, indicating the 
natives of Melanesia, that is, the insular groups (New Britain, 
New Ireland, Solomon, Louisiade, New Hebrides, Loyalty and 
New Caledonia), so named from their "Black" inhabitants 1 . 
Thus the Melanesians are only one section of the group, and as 
they moreover present some special characters, it is in every way 
desirable that they should retain their special name. On the 
other hand no reasonable objection can be made to 
Papuan, which has always been applied by the a P uan8 ' 
Malays to the black aborigines of Malaysia and New Guinea that 
is, to the most typical members of the group and which is more- 
over descriptive of their frizzly " mop-heads," one of the most 

1 Gr. AtlXas, black; ^o-os, island. 




marked physical characters of the race 1 . Where it may be 
necessary to distinguish, the eastern section may be called 



Melanesian Papuans, or simply Melanesians, the western Ma- 
laysian Papuans, New Guinea Papuans, or Papuans proper, as the 
1 Malay *>g (papiiwah), frizzled. "The Malays now understand Papua 


case may be. For the habitat Papuasia seems a convenient and 
appropriate name, analogous to Malaysia, Melanesia, &c. 

At present the Papuan domain is restricted to Melanesia and 
parts of Fiji, practically the whole of New Guinea The Pa uan 
with the neighbouring Torres Strait islands, and domain past 

/i 11 T-^ * -i f an d present. 

most of the smaller groups in East Malaysia as far 

west as Flores inclusive. But in prehistoric times it must have 


also included the whole of Polynesia, as far as Easter Island in the 
extreme east, Hawaii and New Zealand in the extreme north and 
south. This is inferred from the fact that " there are probably few 
if any of the islands of the Pacific in which it [the Papuan element] 
does not form some factor in the composite character of the 

to mean 'frizzled,' as the hair of the Papuans" (F. A. Svvettenham, Malay 
Dictionary, p. 131). The splendid collection of about 600 photographs of 
Oceanic Negroes, published (1894) by A. B. Meyer and R. Parkinson at 
Dresden, is entitled Album von Paptia-Typen, although including great num- 
bers from Melanesia as well as from New Guinea, and this general application 
of the term is steadily growing in favour with ethnologists. The character 
indicated by the word prevails everywhere from Flores to Fiji, and in the 
Solomon group "the whole head of hair has much the appearance of a mop 
placed erect on its handle " (H. B. Guppy, Nature, April 26, 1883). 




natives 1 ." It will be seen that the Papuans must have also most 
probably been the first inhabitants of Australia and Tasmania. 
But whether they had at any time spread over West Malaysia, 
India and Madagascar can no longer be determined, the actual 
relations in these regions being equally explicable by the presence 
either of the Oceanic Negritoes (Malaysia, India), or of the African 


Negroes (Madagascar). De Quatrefages 2 extends their area even 
to the New World, because of the dark colour of the Lower 
Californian aborigines. But colour alone, apart from other 
characters, is not sufficient to determine any racial type, else many 
Semitic Abyssinians and Hamitic Gallas would have to be classed 
as Negroes. The extinct Charruas of South Brazil were also 
described as " black "; but no one has yet spoken of them as 

While agreeing in all essentials with the African, the Papuan 
type presents certain differences, such as more 
fully developed glabella and supraorbital ridges, 
narrower nose, often mesorrhine and prominent, 

1 Flower and Lydekker, p. 748. 
5 Races Humaines t n. p. 406. 

The Papuan 


skull somewhat higher and narrower generally, that is, more 
decidedly hypsistenocephalic. Yet on the other hand dolicho- 
cephaly is certainly a less constant character, so much so that on 
this ground some anthropologists have felt disposed to deny the 
existence of a distinct Papuan type at all. But the variations, 
which may be described as excessive 1 , are obviously due to inter- 
minglings, the foreign elements being in the west the brachy- 
cephalous Malays 2 , in the east the brachycephalous Sawaiori 
(Polynesians). Like their liquid environment, the Oceanic popu- 
lations have always been in a fluctuating state, as sufficiently 
proved by the prodigious expansion of the Malayo-Polynesian 
linguistic family from Madagascar to Easter Island, and from 
Hawaii to New Zealand. 

This stock language has taken exclusive possession of the 
whole area, except Australia, Tasmania, West Papu- 
asia and New Guinea, and even in New Guinea 
some of the coast tribes, such as the Motu of Port 
Moresby, speak pure Malayo-Polynesian dialects. It is as if in 
Africa Bantu were the common speech, not only of the Southern 
but also of the Northern (Sudanese) Negroes, and not only of the 
Negroes, but also of the neighbouring Hottentot, Hamitic and 
Semitic peoples. That the assumed analogy is not strained 
appears from the fact, placed beyond all doubt by 
comparative philology, that the Negroid Malayo- 

Malagasy peoples of Madagascar, the yellow Mon- Polynesian 

goloid Malays of the Eastern Archipelago, some of 

the black Papuans of the same region and of New Guinea, all the 

1 Thus Miklukho-Maclay's measurements show a range of from 62*0 to 86*4 
for the cephalic index of the New Guinea natives, and this observer affirms 
that "we have no right to describe the heads of Melanesians as well as those 
of Papuans as dolicho, but rather as mesocephalic " (Nature^ Nov. 20, 1879). 
Even for the Solomon Islanders Guppy finds a range of from 73 to 84, with a 
mean of 81 for the Treasury natives, and of 74 to 77 for those of St Christoval 
(Nature, April 26, 1883). 

2 This is clearly shown by M. Maclay, who found that for centuries the 
Malays had maintained direct relations with the western parts of New Guinea, 
regularly visiting the Koving coast and other districts for trading purposes, 
and especially to procure slaves for the Sunda Islands (Meinc zweite Excursion 
nach Neue Gttinea, 1874, passim), 




black Melanesians and all the brown Eastern Polynesians as well 
as the mixed Mikronesians, speak idioms belonging to various 
branches of the Malayo-Polynesian stock language. 

The usual explanation of this remarkable phenomenon is that 

^W : : : '''' : -^{f^f9S 


(True Papuan Type.} 

the diffusion of the Oceanic language is due to the migrations of 
the restless and aggressive Malay people, the Orang-laut (" Sea- 


men ") in a pre-eminent sense, who conquered and imposed their 
speech on the surrounding, and mostly inferior, 
insular populations. But the theory, always sus- 

pected because of its simplicity, breaks down com- Polynesian 

r migrations, 

pletely before the facts firmly established by the 
Rev. R. H. Codrington in his classical work on The Melanesian 
Languages^, where it is clearly shown that " Malay is undoubtedly, 
as compared with the languages of Madagascar and the Philippine 
Islands, a simplified form of the common language' 1 (p. 26), and 
that, " as compared with Fijian [a typical Melanesian tongue], the 
languages of Tonga and Samoa [typical Polynesian or Sawaiori 
dialects] are late, simplified and decayed"; in a word that Mela- 
nesian is the most primitive form of the Oceanic stock language. 
It thus becomes self-evident that neither the Malays nor the 
Polynesians, both speaking later dialects, could have diffused this 
archaic form of speech throughout Oceanica. 

Is then a Melanesian to be substituted for a Malay migration 
theory, and is it to be supposed that the admittedly 8tm less to 
inferior race imposed its speech on the more Melanesian 
advanced Malay and Polynesian populations ? Or migra lons * 
is there no solution to a problem in which race and language appear 
to be placed in hopeless antagonism? It has been shown that 
the whole of Polynesia, taken in its widest sense, was originally 
occupied by the black element. It also appears from Mr Sidney 
H. Ray's recent investigations in New Guinea 2 that Motu and the 
other Malayo-Polynesian languages current on the south-east 
coast of that island belong, not to the Polynesian branch, as had 
been supposed, but to the Melanesian, showing later Melanesian 
migrations to that region. This is one of those instances in which 
speech proves to be not merely a useful, but an indispensable 
factor in determining the constituent elements of mixed races. 
But Mr Ray further shows that the languages of the New Guinea 

1 Clarendon Press, Oxford, 1885. 

2 The Languages of British New Guinea^ in Anthrop* Jour. August 1894, 
In this valuable paper the practical identity of the New Guinea Maiva, Motu, 
Loyalupa, Sariba, Awaiama and Dobu with the Melanesian, and especially 
with the Efate of the New Hebrides group, is established on phonetic, structural 
and lexical grounds. 


Papuans proper are fundamentally distinct from the Melanesian 1 , 
consequently from the Malayo-Polynesian ; and that even in 
Melanesia itself there are some dialects, such as the Kiriwina, 
Nada, Misima and Tagula of the Louisiades, which " only partly 
agree with the Melanesian," and which may be regarded as 
possibly "belonging to originally Papuan stocks, upon which 
have been grafted in course of time words and idioms from the 
Melanesian tongues." This " Melano-Papuan " group, as Mr Ray 
calls it, also comprises other somewhat aberrant members of the 
Melanesian branch, such as Alu (Treasury Island), Buka (Bou- 
gainville), Savo (Solomons) and Ambrym (New Hebrides), all of 
which "differ more or less from the typical Melanesian, and 
probably contain some Papuan elements " (p. 1 7). It follows 
that Melanesian is not indigenous in its present home (which also 
includes Mikronesia 2 ), but must have been introduced and im- 
posed upon the Papuan natives by some foreign people in remote 
prehistoric times. This people is none other than the Eastern 
Polynesians, a branch of the Caucasic division, who possibly in 
the Neolithic period migrated from the Asiatic mainland to 
Malaysia and thence eastwards to the remotest islands of the 

Pacific Ocean. The fact that these Polynesians 
explanation. now speak "late, simplified, and decayed " dialects 

of the common Oceanic tongue presents no diffi- 
culty, the explanation being that, while the archaic form was 

1 "They present in nearly every respect the widest possible contrast to the 
Melanesian" (ib. p. 16). 

2 The languages of Mikronesia are undoubtedly Melanesian, but the natives 
are extremely mixed, showing all shades of colour and transitional forms 
between the Papuan, Malay and Polynesian types. In the western groups 
M. Maclay, who visited the archipelagoes in 1876, describes the people as 
nearly akin to the Polynesians, but with a probable Melanesian mixture, 
shown in the curly and even frizzly hair, dark skin and other Papuan 
characters. In the Esheke (Eshikie) group he found the true border-line 
between the frizzly and straight-haired races (Sitzungsberichte der Berliner 
Gesellschaftfur Antkrop., March 3, 1878). Thus the Marianne, Pelew, Marshall 
and Gilbert groups, collectively called Mikronesia, would appear to have 
been originally peopled by Papuans from Melanesia, and to have afterwards 
received numerous colonists both from Polynesia and Malaysia (the Philip- 
pines), besides occasional settlers from Japan and China. See also Dr O. 
Finsch, Reise in der Siidsee, &c., Berlin, 1884, passim}. 


retained or better preserved by the rude Papuan aborigines, it 
became in course of time more "simplified," that is, improved 
amongst the more progressive Malay and Polynesian peoples. 
Compare English with Gothic, and especially modern Danish and 
Swedish with Icelandic. Nobody pretends that the Danes and 
Swedes have derived their " simplified " Norse dialects from the 
archaic Norse still surviving in Iceland, because history tells us 
that it is the other way, that Iceland was colonised from Scandi- 
navia, not Scandinavia from Iceland. In Oceanica common 
sense supplementing the few known facts must supply the place of 
history, and that the above is the true solution of one of the most 
intricate entanglements in the whole range of ethnology has been 
elsewhere more fully explained 1 . Since that explanation was 
given, and questioned because of the " Caucasic factor " introduced 
into the problem, this factor has been accepted by some of the 
foremost living or lately deceased ethnologists. De Qtiatrefages 
amongst others recognises the presence of " the three fundamental 
types in Oceanica 2 ," while Giglioli goes so far as to speak of an 
"Aryan" element in Australia 3 . 

In this Continent, of which Tasmania may be regarded as an 
"ethnical annexe," most anthropologists recognise 
at least two fundamental types beneath a general Han sub-divi- 
physical and linguistic uniformity. That the black Slon 
element forms the substratum is also commonly admitted, and 
may be regarded as self-evident, the colour being often almost 
quite black, while the features and skeletal structure are distinctly 
Negroid. The natives of the Adelaide River (North- West), who 
may be taken as typical Australians, are described 
by a recent observer as " brown-black to almost a 
pure black... the head long and prognathous; eyes 

1 See A. H. Keane, On the Relations of the Indo-Chinese and Inter- 
Oceanic Races and Languages, Jour, Anthrop. Inst. February, 1880. 

2 "Les trois types fondamentaux se retrouvent en Oceanic... Les Blancs 
allophyles [Caucasians] occupent essentiellement la Polyne*sie, les Noirs 
[Negroes] la Me"lanesie...En Malaisie surtout, les Jaunes [Mongols] sont venus 
se joindre aux deux autres types" (Races Humaines t 1889, II. p. 335). 

3 " noto infine che i Tasmaniani erano Negroidi e diversi in razza dagli 
Australian! che io considero Arianoidi degenerati " (Archivio per VAntrop. 
xxiv. 1894). 


deep-set; nasal bones depressed, nostrils large, dilated, and lips 
thick; their legs have practically no calf muscles 1 ." But they 
differ from all other Negro or Negroid races in the character of 
the hair, which is neither woolly nor frizzly, but at most bushy, 
curly or wavy, thick, black, and like the beard (often well deve- 
loped) of somewhat coarse texture 2 . The explanation, suggested 
amongst others by Flower and Lydekker, is that they are probably 
not a homogeneous group at all, as supposed by Huxley, but a 
cross between two already formed stocks. Thus Australia may 
have been " originally peopled with frizzly-haired Melanesians... 
but a strong infusion of some other race, probably a low form of 
Caucasian Melanochroi, such as that which still inhabits the inte- 
rior of the southern parts of India, has spread throughout the land 
from north-west, and produced a modification of the physical 
characters, especially of the hair" (Op. cit. p. 748). It is added, 
however, that the Australians may possibly be mainly sprung 
from a very primitive human type, from which the frizzly-haired 
Negroes may be an offset, frizzly hair being probably a specialisa- 
tion, not the attribute of the common ancestors of the Hominidae 3 . 
Possibly a middle term may be drawn from both of these 
alternatives. The "very primitive human type" is 
** more tnan a m ere hypothesis, as is shown by the 
South Australian tribe presenting Neanderthal cha- 
racters, and inhabiting a district which could easily 

1 P. W. Bassett-Smith, Anthrop. Jour. May 1894, p. 324. With regard 
to the colour of the skin, "infants are a light yellow or brown, but at the age 
of two years they have already assumed the hue of their parents" (Carl 
Lumholtz, Among Cannibals, 1889, p. 132). The same remark is made by 
many other observers. 

2 So Lumholtz : " Hair and beard black as pitch, slightly curly but not 
woolly, seldom straight in the north-east, though straight hair is quite common 
in the rest of Australia, especially in the interior. I only once saw a man with 
his hair standing out in all directions, like that of the Papuans" (Among 
Cannibals^ p. 131). This observer, however, denies the " coarse texture." 

3 These and analogous characters occur elsewhere, as in the north-east, 
where " their projecting jaws make them resemble the apes more than any other 
race, and their foreheads are as a rule very low and receding... the superciliary 
arches very prominent, the cheek-bones high, the temporal fossae very deep, 
nasal bones flat and broad, teeth large and strong" (Lumholtz, ib. p. 260). 


be reached by pliocene man at a time when Australia formed 
almost continous land with the Indo-African Continent. On 
the other hand Australia was equally accessible on the north 
,and north-west sides to primitive migrations both from India and 
Papuasia. That such migrations took place scarcely admits of a 
doubt, and the Rev. John Mathew, who has made a special study 
of this question 1 , concludes that the continent was first occupied 
by a homogeneous branch of the Papuan race either from New 
Guinea or Malaysia, and that these first arrivals, to be regarded as 
the true aborigines, passed into Tasmania, which at that time pro- 
bably formed continuous land with Australia. Thus the now 
extinct Tasmanians would represent the primitive type, which in 
Australia became modified but not effaced by crossing with later 
immigrants chiefly from India. These are identified, as they have 
been by other ethnologists, with the Dravidians, and the writer 
remarks that "although the Australians are still in a state of 
savagery, and the Dravidians of India have been for many ages a 
people civilized in a great measure and possessed of a literature, 
the two peoples are affiliated by deeply-marked characteristics in 
their social system," as shown by the boomerang, which, unless 
locally evolved, must have been introduced from India. But the 
variations in the physical characters of the natives stature 
(5 ft. 4 in. to over 6 ft.), features, muscular development, texture 
of the hair appear too great to be accounted for by a single 
graft j hence Malays also are introduced from the Eastern Archi- 
pelago, which would explain both the straight hair in many 
districts, and a number of pure Malay words in several of the 
native languages', as well as the mental capacity, which is "any- 
thing but despicable." 

Skulls from this region microcephalous, with cephalic index 71, facial angle 68, 
nasal index 53 (very platyrrhine). 

1 Proc. R. Soc. N. S. Wales, xxm. Part 2. 

8 In other respects these languages appear all to belong to an original stock, 
which beyond some dubious verbal resemblances or coincidences, has not yet 
been affiliated to any other linguistic group. Affinities may possibly be dis- 
covered with some of the almost unknown Papuan tongues of New Guinea or 
East Malaysia, most of which appear to agree in their limited arithmetical systems, 
possessing no radicals for more than two, three or at most four. The morpho- 
logy also seems to be somewhat analogous, agglutinating everywhere, with post- 

IQ 2 


All this agrees substantially with Flower and Lydekker's first 
hypothesis, especially if the primitive Dravidians be regarded, not 
as of Mongolia stock, against which there are many objections, 
but as " Caucasian Melanochroi," such as are still represented irr 
Southern India and Ceylon by the shaggy-haired and full-bearded 
Todas and Veddahs. Thus would also be explained the wavy 
hair and thick beard forming the most marked physical trait of the 
Australian aborigines, while the Neanderthal characters persisting 
here and there would be traceable to the Ur-Einwanderung of the 
pliocene precursor from the Indo- Austral Continent. 

This solution of the Australian problem has the advantnge of 
also explaining the position of the Tasmanians, who 
manias 6 TaS " are described b Y Flower and Lydekker (p. 748) a* 
perhaps aberrant Melanesians [Papuans], modified 
not by mixture but by long isolation. The divergence is shown 
especially in the width of the skull in the parietal region, the form 
of the nose, the projection of the mouth, size of teeth and charac- 
ter of the hair. Hence the conclusion of Giglioli amongst others 
that the Tasmanians were " of a different race from the Austra- 
lians," whom "they preceded in the island-continent" (loc. tit.). 
The latter part of this statement agrees with Mr Mathew's view, 
while the supposed racial difference will disappear if the Tas- 
manians be compared, not with the average Australian, but with 
the more primitive groups still surviving in some districts. Thus 
the resemblance amounts almost to identity between the accom- 
panying portrait of a Queensland native from a photograph by 
Mr J. J. Lister, and that of a Tasmanian from a sketch taken in 
the year 1845 by Lieut. F. G. S. de Wesselow, R.N. 

fixes in Australia, with both pre- and postfixes in Papuasia. Thus : smin, man ; 
sniinsi, men ; rosniin t of the man ; rosminsi, of the men. The difference in all 
cases from the Malay o- Polynesian family is fundamental. Thus the phonesis 
is much harsher, and richer in consonantal combinations and sounds, even the 
Australian admitting sibilants, fricatives and aspirates (.r, tk, h), as shown by 
A. B. Meyer and M. Uhle, although denied by Fr. Mtiller (Zur Dippil- 
Sprache in O si- Austr alien > Dresden 1882, pp. 129-30). It should, however, be 
added that A. B. Meyer and G. von der Gabelentz hold Australian to be " im 
geraden Gegensatze zu den melanesischen [papuanischen] " in its phonetic and 
formative systems (Bcitrage zur JKenntniss der melanesischen... Sprachen^ Leipzig, 
1882, p. 384). 




In harmony with this theory is the extremely low grade of 
culture both of these primitive Australian groups 

* " * I asmanian 

and of the Tasmanians, lower even than that of culture eo- 
the European palaeolithic man of the Cheilian age. 
Their rude stone implements have been compared with the 
specimens from Portugal claiming to be of pliocene if not of 
miocene origin. None are ground or polished, or detached from 
the core by pressure, but only by blows in the simplest possible 



way; nor were they mounted on handles, but only grasped in the 
hand, like the eoliths described at "p. 74. This simple art was 
acquired only since the British occupation of the island, so that, 
assuming the accuracy of the accounts, the Tasmanians would 
appear "to have remained to our day living representatives of the 
early Stone Age, left behind in industrial development even by the 
ancient tribes of the Somme and the Ouse...The life of these 
savages proves to be of undeveloped type alike in arts and insti- 
tutions, so much so that the distinction of being the lowest cf 
normal tribes may be claimed for them 1 ." 

1 .E. B. Tylor, On the Tasmanians as representatives of Paleolithic Man> 
Jour. Anthrop. Inst. Nov. 1893, pp. 14952. 


Thus while the lately extinct Adelaide tribe carries us back 
nearly to the Neanderthal physical type, the lately extinct 1 Tas- 
manians recall the mental level of eolithic man in Britain. 

1 According to Mr James Barnard one full-blood Tasmanian, Fanny 
Cochrane Smith, was still living at Port Cygnet in 1889 (Nature, Nov. 14, 
1889). But Mr H. Ling Roth writes that he has obtained three photographs 
of Mrs Smith and some of her hair, showing that "she is not a full-blood 
Tasmanian" (Letter to A. H. Keane, Jan. 20, 1896). 



Asia home of the Mongol race easily accessible to the pliocene precursor 
Transition from the generalised human type to the Mongol variety 
Chief Mongol physical characters Diffusion of the Mongol race Early 
Mongolo-Caucasic interminglings Hence aberrant Mongolic groups 
Mongol Family Tree Chief Mongol sub-divisions Their domain The 
Akkads Early linguistic relations The Mongolo-Tatar sub-division 
Nomenclature : Mongol ; Tatar ; Tiirki Divergent Finno-Turki types 
The Samoyedes The Lapps The Baltic Finns ; Karelians ; Tavastians 
White elements in the Mongolo-Tatar domain Avars Magyars 
Bulgars Osmanli affinities Koreo- Japanese group The Koreans The 
Japanese: Physical qualities; Mental qualities The " Hyperboreans" 
The Chukchi problem The Tibeto- Indo-Chinese sub-division General 
physical uniformity Tibeto- Chinese linguistic relations Function of 
Tone in the Isolating Languages Tibetan linguistic affinities Indo- 
Oceanic linguistic relations The Indonesians The Malay problem 
Malay physical type Malagasy affinities Malayo- Polynesian linguistic 
relations Ethnical relations in the Philippine Islands. 

IT is admitted by all ethnologists that Asia is the original home 
of the Mongolic division, a fact which harmonises Asia home 
well with the view that the vanished Indo- African of the Mongol 
Continent was the cradle of mankind. From that race ' 
region the pliocene precursor had easy access through India 
itself to the Central Asian plains and plateaux. At present the 
peninsula appears cut off on all sides by lofty ranges from the 
mainland, although recent military surveys have revealed a con- 
siderable number of relatively easy passes, giving access through 
the Soleiman Mountains to the Iranian tableland. But in the 
pliocene epoch all these ranges stood at a much lower level than 
they now do. Both the Arakan-Yoma, now blocking the way to 
Indo-China, and the Sivalik foothills, date only from the latter 


part of the tertiary era, when the Western Himalayas themselves 
ui -. were probably not more than 20,000 feet high, or 

accessible to r J o 

the Pliocene nearly 10,000 less than at present. Even the 
precursor. Tibetan plateau, now the highest on the globe, was 

a marine bed in the cretaceous age, since when it has been slowly 
raised to its present level. In general "the extra-peninsular 
ranges, the great Indo-Gangetic plain, the northern margin of the 
peninsula, and the western coast owe their origin to another great 
series of earth-movements which took place during the tertiary 
era 1 /' Consequently the way was open from India to the very 
heart of the continent, that is, to the shores of the then flooded 
Central Asian depression, at the very time when pliocene man 
began to spread northwards from the Indo-Austral regions. 

Such a precursor, migrating northwards to a new environment 
on the Central Asian plateau, as at that time consti- 

Transition , . , , . . 

from the gene- tuted, might pass by easy transitions to a form ap- 
" proximately like that of the ideal Homo Mongolians 

Mongol described in Chap. X. Neither colour of the skin, 

variety. *. 

texture of the hair, nor stature could present any 

difficulty, for in all these respects the Mongol type stands actually 
nearer than does the Negro to that of the precursor as conceived 
by de Quatrefages. Hence the unsatisfactory nature of all 
attempts made to derive the yellow and white varieties from 
the black, which is generally but wrongly assumed to be in all 
particulars the best representative of primitive man. It is mainly 
in the form of the skull, its extreme prognathism and dolicho- 
cephaly, as well as in the disproportionate length of the arms and 
slight muscular development of the calves, that the Negro stands 
nearest to the anthropoids. Perhaps in most other respects the 
Mongol takes this position, although Topinard has noticed that 
in the texture of the hair the white comes nearest to the apes, the 
black differing most, while the yellow is intermediate. Such 
results should be expected on the theory here assumed of inde- 
pendent ascent from the prototype, whereas they would be 
inexplicable on the opposite assumption of successive transitions 
from one human variety to another. 

1 R. D. Oldham, The Evolution of Indian Geography, in Geograph. Jour. 
March, 1894, p. 180. 


Taken as a whole, the typical Mongol differs from the other 
divisions mainly in the general yellowish colour of chief Mongol 
the skin, the broad flat features, with very prominent physical 
anteriorly projecting cheek bones, small mesorrhine 
nose, mesognathous jaws, brachycephalous head, slightly developed 
superciliary ridges and glabella, somewhat sunken eyes with 
narrow almond-shaped aperture between the lids, a vertical fold 
of skin over the inner canthus and outer angle slightly elevated. 
This oblique eye with its "third lid" is a highly characteristic 
trait, constant in the more typical groups, and exclusively found 
in the Mongol division. The black, lank and rather coarse hair, 
almost if not quite circular in transverse section, is also a constant 
but not an exclusive character, being equally common to the 
American division, and forming the most marked physical link 
between Homo Mongolicus and Homo Americanus. It seems to 
justify the assumption of an original generalised Mongolo- American 
type, from which the American branched off at an early date prior 
to later differentiations, as represented in the Family Tree of the 
Hominidse (p. 224). 

After the separation the parent stem continued to spread over 
a great part of the continent, reaching its extreme Diffusion of 
eastern limits probably in the palaeolithic age, the Mongol 
passing later southwards into Malaysia, and pene- 
trating in neolithic times into Europe, but apparently not into 
Africa. This early expansion of the Mongol race, of which there 
is monumental evidence in Mesopotamia, and abundant ethnical 
proof in Indo-China and the Amur basin, brought about fresh 
groupings and interminglings, not only with kindred .tribes, but 
also with Caucasic peoples, who had already in remote times 
spread from their primeval homes in North Africa and Europe 
eastwards to Japan, south-eastwards to India and Indo-China, 
and thence to Malaysia, Australasia and Polynesia. Thus arose, 
not only on the confines but in the very heart of the Mongol 
domain, those Mongoloid and Caucasoid aberrant groups, such as 
the Malaysian Indonesians, the Mesopotamian Ak- 
kads, the Dravidians of the Indian peninsula, the 

Ugrian Finns, and the Ttirki peoples, wrongly called in * cr - . 
Tatars, all of whom are found fully constituted long 







(Full face.) 


( West Mongol Type.) 



before the dawn of history, but whose ethnical affinities have 
remained an unsolved problem. But, speaking broadly, it may 
be confidently said that the explanation of these ethnical puzzles 
will be found in the frank recognition of Mongolic and Caucasic 
elements interpenetrating each other at various points of their 
respective territories from the earliest times 1 . In the presence of 
distinctly fair types and regular "European" features in Man- 
churia, Korea, Yezo, Turkestan, parts of Siberia, and Malaysia, 
the assumption must be abandoned that these regions have always 
been the exclusive appanage of the yellow race. In the chapter 
devoted to the Caucasic division, this important subject will have 
to be dealt with more fully. Meanwhile it will suffice here to 
point out that in the accompanying Family Tree of the Mongolic 
division all those aberrant groups find a place, which can be 
shown to belong fundamentally to the Mongol stock. Here 
language becomes an important factor, to which 
appeal may be made in doubtful cases, while historic 
evidence is available in determining the constituents 
of Magyars, Bulgars, Uzbegs and some other later groupings. 

Here, as in the Negro Tree, two great limbs are seen to branch 
off from the parent stem nearly simultaneously 
the Mongolo- Tatar to the left and the Tibeto-Indo- 
Chinese to the right. From each of these springs 
an extra-continental branch, the Mongolo-Tatar passing with the 
Eskimo to America, the Tibeto-Indo-Chinese with the Oceanic 
Mongols to the neighbouring Indo-Pacific waters. Thenceforth a 
relatively close connection is maintained between the various 

1 On this point Dr Hamy aptly remarks: "Nous passons d'une race a 
1'autre par des transitions insensibles, et nous avons pu ainsi nous rendre 
compte de Textreme difficult^ que pre*sente habituellement la delimitation 
scientifique des Jaunes et des Blancs. Les vallees de la Siberia occidentale 
sont parcourues dans les hautes latitudes par des peuples, com me les SamoSedes, 
Kanirs et autres, chez lesquels les variations individuelles sont vraiment fort 
e*tendues et conduisent, a peu pres sans hiatus, du Mongol au Lapon. Ailleurs, 
dans les mmes zones, les types intermediaires e"tablissent d'autres passages 
presque insensibles du Lapon au Finnois et du Finnois an Slave. On peut ainsi 
aligner des series d'observations continues entre les plus exageres des Jaunes, 
et certains Blancs tout a fait aveVes." (Les Races Jaunes in U Anthropologic, 
May-June, 1895, p. 249.) 



Tibeto-Indo-Chinese sub-branches, whereas the Mongolo-Tatar 
ramifications are not only far more numerous, but also develop 
more independent secondary branches, such as the Akkad, the 
Koreo-Japanese, the Finno-Tatar and the Mongol proper, some 
confined to the Asiatic mainland, others spreading eastwards to 
the Japanese archipelago, or westwards far into Europe. Europe, 
however, may be regarded as to some extent an ethnological, as it 
is altogether a geographical, dependency of Asia. Thus the 
whole of the northern section of the eastern hemisphere is seen to 
be largely occupied by Mongol or Mongoloid peoples, the regions 
from which they are either partly or altogether excluded being 
India, Irania, Arabia, North Africa and West Europe. Such 
throughout the historic period has been the domain of the Mongol 
division, here and there modified from time to time by the vicissi- 
tudes of the secular struggle for ascendency maintained with the 
conterminous Caucasic populations. 

For the Mongol division the historic period dates from the 


earliest records of the Mesopotamian Akkads, founders of the 

oldest known civilization in Babylonia. Here some 

of the figures brought to light by M. de Sarzec at 

Tell-Loh, site of the ancient city of La gash (4000 2500 B.C.), 


show distinctly Mongolic traits in the prominent cheek bones, 
oblique eyes and generally flat features, standing out in marked 
contrast to the almost pure Semitic type of the later Assyrian 
epoch 1 . We have also seen (Ch. IX.) that the Akkad language 
shows some striking resemblances to Chinese, while most Akkad 
students are rather inclined to affiliate it either to the Finnic or to 
the Ttirki branch of the Mongolo-Tatar linguistic family. The 
reconciliation of these apparently contradictory views may be 
found in the now established fact (p. 207) that Chinese was itself 
formerly polysyllabic, and may consequently have sprung from a 
common Tibeto-Mongol form of speech, of which Akkad is the 
earliest and nearest representative. If continued in this direction, 
Akkad studies may lead to a satisfactory solution of the Tibeto- 
Mongol problem, and to the recovery of the primordial unity of 
the Mongolic division, which appears in our Mongol Family Tree 
to be split from prehistoric times into two great subdivisions. 

The expression "Mongolo-Tatar" applied to one of these 

The Mon- subdivisions, is perhaps too firmly established to 

goio-Tatar be now set aside. But "Mongolo-Ttirkic" would 

subdivision. L . , , f ,, - , ... 

certainly be preferable, though still not quite satis- 
factory, as seeming to exclude the Eskimo, the Koreo-Japanese 
and the Finnic groups. The expression Ural-Altaic has in recent 
years come into favour as a convenient alternative and is certainly 
better than the misused and discredited "Turanian 2 ." It is of 

1 On Plate xxv. of de Sarzec's Dlcouvertes tn Chaldle " a small head is 
figured in which an obliqueness of the eyes is clearly noticeable" (T. G. Pinches, 
Types of the Early Inhabitants of Mesopotamia, in Jour. Anthrop. Inst. 1891, 
p. 99). The head of No. i, Plate XII. figured and restored by Mr Pinches 
(pp. 87, 88), also shows a general Mongolic expression in the flat face and 
well-marked malar bones flattened in front. No importance can be attached 
to Mr Pinches* restoration of the mutilated nose, which was probably smaller 
than here shown. 

2 Whether containing the root of Tilrk, or, as seems more probable, trace- 
able to an Aryan word meaning "swift" (cf. Skt. tvard= haste, speed), Tdra 
(later Tiirdri) was applied in the early Persian records to the region north of 
Ariana (East Irania) now known as Tiirkistan, " Land of the Turk." Thus 
was indicated the sharp contrast, the everlasting antagonism, between the 
settled Aryas (root ar t to ear, to plough) and the nomad Tiiranl> swift-moving 
predatory hordes then as now. When comparative philology began to extend 


course geographical, and like the analogous " Indo-European " 
is defective, excluding large areas occupied by Mongoloid peoples 
of Mongolo-Tatar speech. Here it will be used like Schott's 
Ugro-Altaic in the widest sense, so as to embrace the whole 
region from Lapland to Japan inclusive, a region occupied by 
peoples of more or less homogeneous physical type, and now 
shown to speak idioms ultimately traceable to a common stock 

The great objection to "Mongolo-Tatar" lies in the fact that it 
is tautological, both terms of the compound form being historically 
referable to Mongol peoples proper, so that Tatar is wrongly 
taken as synonymous with Turki. No objection 
can be made to the first component, whether 
derived from mong, "brave," or from the Mongol 
tribe of which Jenghiz-Khan was chief, and which in the i2th 
century was seated near the Kara- Kara mountains north of the 
Gobi Desert. But Tatar (plural of Tata) was never the name of 
any section of the Tiirki branch, to which it is now collectively 
applied. It appears to be a Tungus or Manchu word, meaning 
either "archer" or "nomad," and first occurring in Chinese records 
of the 9th century in reference to certain Mongol tribes which 
were later driven by the Khitans southwards to the In-Shan moun- 
tains about the great bend of the Hoang-ho river. Here the 
predatory Mongols and Tatars, all closely related members of the 
Mongol group proper, were welded into one nation by Jenghiz- 
Khan, a Mongol on his father's side, and a Tata on his mother's. 
That Tatar became dominant in the west was largely due to the 
fact that the Tatas generally formed the van of the Mongol expe- 
ditions westwards. At an early date Tatar took the form Tartar 
by association with the Tartarus of classic mythology, as in the 

its sphere from the Indo-European to the Central Asiatic linguistic domain, 
Turdn naturally supplied the comprehensive term "Turanian'* needed to 
distinguish the Mongolo-Tatar from the Aryan linguistic family. But while 
Aryan has held its ground, Turanian has fallen into abeyance, thanks to its 
misuse by popular ethnographists, who made it a convenient receptacle for 
almost everything non-Aryan in the Eastern, and even occasionally in the 
Western, Hemisphere. At present Turanian is the shibboleth of unscientific 
and inaccurate writers on ethnological subjects. 


letter (1241) of Louis IX. to Queen Blanche 1 . Thus it happened 
that Tatar or Tartar, originally the name of a Mongolian tribe, 
was gradually transferred to the western group whose proper name 
always has been and still is Tiirkt, though in many places ruled 
by Khans of real or pretended Mongol descent. The powerful 
Kipchak empire, founded by Batu-Khan, grandson of Jenghiz, 
was mainly inhabited by Kumans, Pechenegs, and other Ttirki 
peoples, and when the empire was broken into fragments, each 
section still continued to be ruled by Tatar Khans and to be 
called a Tatar Khanate. Thus originated the expressions "Sibe- 
rian Tatars,' 7 Kazan, Astrakhan, Krim (Crimean) and other Tatars, 
that is, Tiirki peoples ruled by Tatar princes of Jenghiz- Khan's 
dynasty. But the peoples themselves have always disclaimed the 
name of Tatar, calling themselves and their language Tiirki. 

This word is of far more venerable origin than the Mongol 
term which has partly usurped its place. It is traceable in its 
mutilated Chinese form Tu-kiu back to the 2nd century B.C. when 
a people of that name dwelt in the Altai region. Here they 
gradually rose to great power, and in the first century of the new 
era their name had already reached Europe, the Turcce being 
mentioned both by Pomponius Mela (i. 22) and by Pliny (vi. y) 3 . 
The Hiung-nu and the On-Uighurs, founders of vast but unstable 
empires, were all of Tiirki stock, as were also the bulk of Attila's 
hordes, that is, "the Huns whom we commonly call Turks" 
(G. Theophanes, 8th century). In 569 Sinjibu, Kha-Khan 
("Great Khan 7 ') of the Altai Turks, received an embassy from 
Justin II. of Constantinople, and ever since that time the Turks, 

1 "Erigat nos, Mater, aeleste solatium, qiiia si perveniant ipsi, vel nos 
ipsos quos vocamus Tartaros ad suas tartareas sedes retrudemus, vel ipsi nos 
omnes ad cselum subvehent." But it would almost seem from this text and 
from other circumstances as if the form Tartar had already been established, 
and the word occurs in fact in earlier documents (1237 and 1240). A vast 
amount of information on the early history and relations of the Mongolo-Tiirki 
peoples is embedded in Sir H. H. Howorth's monumental but ill-digested 
History of the Mongols, 1876-80. 

2 W. Thomsen reads the name Tiiirk in the scarcely decipherable rock 
inscriptions of the Yenisei, which he refers to a Tnrki dynasty ruling in that 
region in the 8th century (Paper submitted to the R. Academy of Denmark, 
Jan. 1894). 


under one name or another, have maintained almost uninterrupted 
relations, hostile or friendly, with the nations of the West, over- 
throwing the Byzantine Empire (1453), penetrating up the Danube 
to the gates of Vienna (1683) an d still holding their ground in the 
Balkan Peninsula, Anatolia and parts of Irania. 

The Tiirki type, originally Mongolic, had at an early period 
been profoundly modified in many places, and D i ver ent 
especially in the north by contact with peoples of Finno-Turki 
Caucasic race, whence the frequent mention of ypcs ' 
"red hair," "green eyes," and "white complexion" in the early 
Chinese records. By some ethnologists these modifications have 
been attributed to interminglings with Ugrian (Siberian) Finns, 
the Finnish race itself being regarded as originally not Mongolic 
but Caucasic. De Quatrefages habitually calls them "white 
Allophylians." Many of the European Finns 1 , and especially the 
Baltic group, have undoubtedly been largely assimilated to the 
surrounding populations, although even these retain certain 
physical and mental characters, such as peaky eyes, somewhat 
flat face, round head, dull sullen temperament, which, combined 
with their pure Ural-Altaic speech, betray their primordial Mongol 
affinities. These affinities become more marked in the direction 
from west to east, until the Samoyedes, Soyotes and other Finns 
of Siberia, true home of the race, show all the characters of the 
Mongolic type often in an exaggerated form. 

Thus the Samoyedes 8 of the Ob basin, studied by Castren, 

1 The, European Finns call themselves Suomilaiset, usually interpreted 
"Fen People," from suo, fen, swamp; but this cannot be because the m of 
suomi is radical ; hence the assumption that " Finn " is a Teutonic translation 
of Suomi, in the sense of "Fen People," also falls through. Moreover to 
derive this word from Old Norse fen is philologically impossible. Here the 
e arises by umlattt from an original a, as in the Gothic /ant, whereas Finn 
(Fenni, Finni y Tacitus, Germania 46; Pliny 4. 13) goes back to a time 
long prior to the appearance of umlaut in the Teutonic languages. See W. 
Thomsen, Ueber den Einflnss der germanischen Sprachen auf die finnisch- 
lappischen^ Halle, 1870, note p. 14. 

2 Properly Hasovd or Nentzi, both terms meaning "men," and current, the 
former chiefly in the Ob basin, the latter west of the Pechora river. The 
word Samoyede, for which the absurd popular etymology " Self-eaters," in the 
sense of Cannibals, has been found in the Russian language, appears to be also 

K. 20 


Finsch and especially Stephen Sommier, have the 
Sa^oyedes. true Mongolic eye, fold and all, the character- 
istic short nose, prominent malar bones, flat 
features, low stature scarcely exceeding 5 ft. i or 2 in., lank 
dark hair ranging from a deep chestnut to black, and distinctly 
round head (Sommier's mean cephalic index 84-44) * Although 
their southern neighbours, the Ugrian Voguls and Ostyaks, present 
some marked differences, especially in their lighter complexion, 


(North Mongolic Type.) 

chestnut and even blonde hair, and long head (index 79*28), the 
Mongol type is conspicuous enough in their flat features, small 
nose, slightly oblique eyes, and short stature (little over one per 
cent, taller than the Samoyede). The hair is described as "red" 

a national designation, in which Samo is to be equated with the Finnish 
Suomi and the Lapp SamJ, as in Suomilaistt, Samtiats (A. H. Keane, The 
Lapps, p. 3). According to Mr Fr. G. Jackson the present pronunciation is 
Samoyad, or even Samyad, at least in the districts visited by him during his 
expedition to Waigatz Island: "Mr Jackson found that the Yuraks of the 
Trans-Pechora country invariably pronounced the name as if it were Samo-yad, 
or even Sam-yad" (The Great Frozen Land, 1895, note by Mr Arthur Monte- 
fiore, p. 54). So also Mr Trevor-Battye, Ice-bound on Kolguev, 1895, p. xxii. 
1 Sirieni 9 Ostiacchi e Samoiedi delf Ob, Florence 1887, p. 150. 


by Topinard and other ethnologists; but this is a mistake arising 
out of a faulty translation of the original account 

. The Lapps. 

given by Pallas, which was itself inaccurate, but 
which has now been rectified by Sommier (p. 63). It is note- 
worthy that the hair of the Lapps, most aberrant of all the 
western Mongoloid peoples, is at present generally brown or light 
chestnut, whereas in Linnets time it was normally black 1 . This 
rapid modification of a marked physical character, attributed by 
the writer 2 doubtfully to crossing with the fair-haired Scandinavians, 
is shown by Sommier 3 to be more probably due to alliances with 
the blond Quaens, that is, Finnish immigrants into Lapland. But 
whatever be the cause, the fact is important, as illustrating the 
analogous changes by which in the course of ages the Baltic Finns 
themselves have been largely assimilated in appear- 
ance to the average European. In this group F 
Retzius 4 distinguishes two well-marked types, the 
eastern Karelians, tall, slim figures with regular features, straight 
grey eyes, brown complexion, and chestnut hair hanging in ringlets 
down to the shoulders, and the western Tavastians, the "white- 
eyed Chudes" of the Russians, broad thick-set figures, small and 
slightly oblique blue eyes, light flaxen or towy hair, and white 
complexion. It would almost seem as if the Tavastians were the 
issue of a German graft on a Mongol stock, while the Karelians 
represented a Slavo-Mongol mixture in which the original Mongol 
element was largely eliminated. In this respect the Karelians 
resemble the more easterly Permian Finns, and especially the 
Siryanian group, who dwell on both sides of the northern Urals, 
and who are distinguished from the neighbouring Samoyedes by 
their white colour, blonde or light chestnut hair, large brown or 
grey eyes, and straight nose. Some of these Russified Finns have 

1 " Cap} His nigris, brevibtut, rectis" Sy sterna Natura. 

2 The Lapps, 1885, p. 7. 

3 Sttt Lapponi e sui Finlandesi Settentrionali, Archivio per FAntropologia, 
xvi. 1886, p. 162. The general Mongolic affinities of the Lapps, affirmed by 
the writer (op. cit. passim), but denied by many anthropologists, is accepted by 
this observer: " Egli ammette che i Lapponi sono di origine Mongolica, 
ed in questo, dando al termine Mongolico un senso molto largo, andiamo 
d' accordo" (pp. cit. p. 52). 

4 finska Kramer, passim. 

20 2 


even developed a full beard, while others still betray their Mongol 
descent in their broad heavy features, small nose, and large malar 
bones, "recalling the Tavastian type of Finns'." 

It is generally supposed that the region about the headwaters 
of the Yenisei was the original home of the Finnish 
race anc * ^ ere st *^ surv i ye a ^ ew isolated Samoyede 

Mongoio- tribes, such as the Koibals, Karagasses, Kamassintzi 


and Soyotes. Although some of these have inter- 
mingled with the neighbouring Tiirki tribes and now speak Ttirki 
dialects, the Turks of Central Asia could not have acquired their 
Caucasic features from this source, the affinities of the Upper 
Yenisei Finns being, not with the Slavonised or Germanised 
western groups, but with the North Siberian Samoyedes of pro- 
nounced Mongol type. The same Caucasic strain has moreover 
been traced through Manchuria and Korea to Japan, regions 
where no Finns have ever been heard of, but where a distinctly 
Caucasic element still survives in the Ainu aborigines of Yezo. 
The early Chinese records have preserved the memory of other 
"Allophylian Whites," such as the Wusuns, an extinct historical 
nation of Central Mongolia described in the annals of the Han 
dynasty as a tall fair race, with red hair and green eyes, who were 
gradually driven by the Mongols westwards to the Tarim basin 
(Kashgaria). Thus the Mongolo-Tatar populations are everywhere 
found from remote prehistoric times interpenetrated by primitive 
Caucasic peoples, and it is to the interminglings of these two 
elements that must be attributed the Caucasic characters noticed 
in all ages amongst the Finno-Tatars and their more remote allies 
of Manchuria, Korea and the Liu-Kiu (Lii-Chii) Archipelago. 

During their later migrations southwards and westwards the 

Finno-Tatar peoples (Avars, Magyars, Bulgars, 

Osmanli and other Turks), underwent still more 

profound transformations. Nearly all became assimilated by misce- 

genation to the Caucasic type, some (Avars) being completely 

absorbed, others (Bulgars) retaining but slight traces of their 

Mongol descent and nothing of their Finno-Tatar speech, others 

again (Magyars, Osmanli) losing their physical characters, but 

1 Sommier, Sirieni, &c,, p. 66. 


preserving their highly agglutinating Ural-Altaic languages l . The 
Avars, whose very name has perished, unless they are still repre- 
sented by a small group of wild hillmen bearing that designation in<^ 
the Caucasus, were formerly dominant from the Don to thj| 
Middle Danube, where they clashed swords with the legions JP 
Charlemagne. In Pannonia (Hungary) they were replaced by 
the kindred Magyars of Finno-Tilrki stock, who in the Qth century 
crossed the Carpathians and pitched their tents on the banks of 
the Theiss and Danube, where they are still the dominant people. 
The Magyars are ethnologically an extremely inter- 
esting nation, who for about a thousand years have 
preserved their Finno-Ttirki speech intact amid a congeries of 
Aryan-speaking populations, while in their new environment their 
Mongolic physical type has gradually conformed to the normal 
European standard, perhaps partly by convergence, but mainly 
by continuous crossings with their German, Slav and Rumanian 
neighbours. Mentally also the evolution is complete, and the 
frank, chivalrous, intelligent and highly cultured Magyars of the 
present dny differ as much from their rude nomad forefathers 
roaming the northern steppe as does the present imperial race of 
Englishmen from their Romano-British forerunners. But how are 
such a people to be classified? No doubt some of the peasantry, 
and especially the so-called Szeklers* of Transylvania, are still dis- 
tinguished by somewhat coarse Mongoloid features, whether 
inherited or acquired by fusion with the Avars. But were it not 
for their Ural-Altaic speech, the most experienced anthropologist 
would fail to detect a drop of Mongol blood in the regular, often 
handsome features, white skin, shapely pliant figure and quick 
flashing glance of the average Magyar of the present generation. 
It thus becomes evident that, when the details are reached, all 
classifications resolve themselves into more or less convenient 
groupings of the transitional forms by which the primary divisions 
are everywhere connected on their ethnical borderlands. 

So with the Bulgars, a horde of Volga 8 Finns, who in the 

1 For an explanation of these phenomena see Chap. IX. 

2 Properly Szekely, "Borderers." 

3 The Byzantine chronicler Nicephoras Gregoras (i4th century) states 
expressly that the Bulgars took their name from the Bulga (Volga), which 





(Profile.) (Full face.) 



(Turki Type,) 


7th century (678) moved southward and settled in 
Moesia on the right bank of the Lower Danube. 
Here their fate was somewhat different from that of their remote 
Magyar kinsmen. Failing to preserve their ethnical indepen- 
dence, they had already in the xoth century exchanged their 
Finnish speech for a Slav tongue, and were thenceforth classed 
with the Slavonic branch of the Caucasic peoples. Yet the 
Mongol physical characters were never quite eliminated, and are 
still perceptible in the somewhat broad flattish face, long black 
hair, small slant eyes, heavy figures, and rather sluggish tempera- 
ment of the modern Bulgarians. In their original home, "Great 
Bulgaria," the Mordvins, Cheremisses, Chuvashes, Votyaks and 
other kindred Volga Finns, are also being slowly Slavonised, and 
on linguistic maps already appear like so many ethnical islets lost 
amid the vast sea of surging Russian nationality. A similar fate is 
overtaking the Bashkirs, Tepyaks and other Finno-Tatar groups of 
the Southern Urals, as well as the Baltic Finns south of the Gulf 
of Finland, where those of Kurland and Livonia have already 
disappeared; none of Finnish speech now survive in this region 
except the historical Esthonians, of whom King Alfred has left an 
interesting account in his translation of Orostus. 

Like the Finns, the European Turks (Osmanli) have lost most 
of their Mongol physical characters 1 . But while 
the Finns have generally become xanthochroi 
(blondes), the Turks have approximated more to 

traverses their country, "Great Bulgaria," so called in contradistinction to 
" Little Bulgaria " (Moesia) south of the Danube. 

1 These characters are thus described by Dr Hamy: "Aplatissement 
parieto-occipital, commun & tons les Turcs. Ce trait signale*tique tres habitual, 
tres manifesto, permet deja d'etablir entre le Mongol et le Turc une difference 
imme'diatement appreciable. II en est un second plus frappant encore et qui, 
combine* avcc le premier, donne a la botte cranienne du Turc, qu'il soit 
Yakoute ou Turcoman, un aspect cubdide. C'est la tendance de la tete a se 
developper en hauteur, juste en sens inverse, par consequent, de Taplatisse- 
ment vertical du Mongol. La tte du Turc est done a la fois plus haute et 
plus courte ; elle est aussi un peu moins large a proportion et Pindice cepha- 
lique est seulement sous-brachycephale. La face, s'harmonisant, comme il 
convient, avec le crane ainsi quelque peu re*treci, est moins e*panouie; par 
contre le squeletle nasal s'accentue plus encore chez le Turc que chez le 




the melanochroic (dark) subdivision of the Caucasic type, a differ- 
ence readily explained by their exogamous alliances with the 
Circassians, Abkhasians and other Mohammadan peoples of the 
Caucasus. Their Mongol descent is now chiefly shown by their 
Tilrki language, which despite the Arabic and Persian forms of 
the literary standard, still preserves the peculiar agglutinating 
structure of Ural-Altaic speech. When they are followed along 
the line of their westerly migrations to their Central Asiatic homes, 
the Turks, like the Finns, are also seen to gradually approximate 
to the Mongolic physical type. This may be seen by a detailed 
study of the Anatolian Turks, the Kizil-Bnshes of the same region, 
the Afars, Qa-jars and other Tiirki nomads of Persia, the Turko- 


(Turko- Iranian Mixed Type.) 


(Manchu Type.} 

mans of Western Turkistan, the Uzbegs, Kara-Kalpaks and 
others of the Oxus basin, the Kirghiz hordes of the West Siberian 
steppe, the East Siberian Yakuts, the Solons of the Amur basin 

Mongol, et vous avez pu voir sur des Ansariehs, par exemple, des cas de 
macrorhinie veritablement surprenants" (V Anthropologist 1895, May-June, 
p. i 4 ). 


and the Hor-Soks of the Tibetan plateau. The Tilrki branch 
proper is thus found to cut obliquely across the heart of the 
continent from the Lena and Amur basins to the Bosphorus, 
interrupted here and there by Mongol, Iranian, and other elements, 
but everywhere showing remarkable linguistic uniformity amid all 
the transitions between the Mongolic and Caucasic physical types. 
From the Mongolo-Tatar bough of the Family Tree a slender 
branch comprising the Koreo-Japanese group is 
seen to ramify independently between the two 
main subdivisions of the parent stem. The rela- 
tive position of this group to the other members of the family, long 
a subject of discussion, may now be regarded as settled. That the 
separation took place at a very remote period is evident from the 
difficulty observers have had in recognising the connection, which 
even now has been established perhaps as much by the aid of 
language as of physical characters. This may be explained by the 
fact that both the Koreans and Japanese are mixed peoples, 
yellow and white elements prevailing in the former, yellow, white 
and perhaps brown in the latter, whereas their languages are 
unmixed and fundamentally related with the Ural-Altaic family. 
Korean was first shown by Mr W. G. Ashton to be remotely con- 
nected in its verbal and structural character and phonetics with 
Japanese. "It seems probable that the distance which separates 
Japanese from Korean... is not greater than that which lies 
between English and Sanskrit.... Everything considered, we may 
regard them as equally closely allied with the most remotely con- 
nected members of the Aryan family 1 /' Since then a distant 
relation has also been established between Japanese and the other 
branches of the Ural-Altaic stock language. But here the affinity 
is exceedingly faint, less even than that now established between 
the Hamitic and Semitic groups. It would appear however to be 
of a fundamental nature, due, not to later contact, but to common 
descent. On this obscure philological problem, which so nearly 
concerns Ural-Altaic ethnical affinities, much light has been 
thrown by Dr Heinrich Winkler in his scholarly treatise Japaner 
und Altaier (Berlin, 1894). Here it is shown that all the essential 

1 A Comparative Study of the Japanese and Korean Languages^ in Jour. 
JR. As. Soc. 1879, p. 360 et seq. 


fentures of Japanese, and consequently also of Korean, find their 
counterpart in the Finno-Ugrian group. Numerous identities 
have also been traced between the radical elements and primitive 
vocabularies of both families, so that little doubt now remains of 
their fundamental unity. In this respect Japanese and Korean 
would appear to stand in much the same relation to Finno-Tatar 
that the Hamitic does to the Semitic linguistic family. In both 
cases no doubt the disparity is enormous, and such wide diver- 
gences from original stock languages must have taken a vast 
period of time to accomplish. But such a consideration can 
have weight only with those who, despite accumulating evidence 
of great antiquity, still persist in limiting the existence of man to 
a few thousand years. 

Their common descent is also clearly perceptible in some of 
the physical features of the Koreo-Japanese groups. 

The Koreans. ,. r ' , , ... . . r 

I he Koreans, who take an intermediate position 
between the continental and insular Mongoloid peoples, are 
somewhat taller and more robust, with much lighter complexion 
and far more regular features than the average Mongol. As 
amongst the neighbouring Manchus, greenish, grey and even blue 
eyes are not uncommon, and the fusion of yellow and white 
elements is perhaps more marked than elsewhere in north-east 
Asia. Ernst Oppert 1 everywhere met people, and especially 
children, with such regular features, florid complexion, light hair 
and blue eyes, that they could scarcely be distinguished from 
Europeans. The national records speak of two primitive races, 
the Sien-pi and San-San, apparently representing yellow and white 
types, who were gradually merged in the present Kao-ri (Kao-li, 

Their Japanese neighbours are the outcome of more complex 

interminglings 2 . According to the national tradi- 
japanese. ^ ons tnev arr ivcd from the south and south-west, 

and gradually spread over the archipelago, driving 
the Caucasic Ainu aborigines northwards to Yezo, and no doubt 

1 Reisen nach Korea, Leipzig, 1881, passim. 

2 "Among the Japanese there are three distinct types noticeable in the 
living subject (Rosny), and a fourth which we may gather from an examination 
of skulls " (Topinard, Anthropology, p. 445). 



here and there mixing with them, though nowhere to any con 
siderable extent. Some appear to have arrived from the southerr 


Malay lands (Formosa, the Philippines), while others may hav< 
come from Polynesia. But there is nowhere any evidence of th( 


black or Negrito element that has been spoken of, and all the 
evidence points to Korea as the original home of the great 


majority, and especially of the dominant classes. Amid much 
diversity, all these elements have merged in a marked Japanese 
type, which can generally be recognised, and which 
* s characterised by a flat forehead, great distance 
between the eyebrows, small but well-formed nose, 
with slightly raised nostrils, no glabella, nor any depression at the 
root of the nose, small black eyes, rather less oblique than the 
Chinese, lank black hair, scant or no beard, disproportionately 
short legs and low stature, shown by the measurements now 
taken of the conscripts for the army to average about 5 ft. 4 or 
5 inches. The complexion is sallow, or dirty olive-yellow, but 
" it is curious how the (face) complexion of these people differs 
from the body complexion. In the course of two visits to Japan, 
in which I travelled much in various parts of the country, I saw 
many hundreds of naked Japanese, the bathing of both sexes in 
company being at that time the rule, and I was struck particularly 
with the fact that, in spite of their sallow or yellowish com- 
plexions, their bodies were whiter than those of Englishmen or 
even English women. The Chinaman, however, strips yellowish 1 ? 
u The Koreans are notably taller than the Japanese; and it is 
on the islands of Tsushima and Iki, in which Korean blood 
predominates, that the height of the men averages one inch more 
than on the main island, Hondo 2 ." 

It was seen (p. 45) that there is no necessary correlation 
between physical qualities and mental endowment. 
iHentaiquah- ,py g et hnological datum is perhaps better illustrated 
in the Japanese than in any other race. Compared 
with the average Chinese and especially with the Manchus and 
Koreans, they are but a feeble folk, no doubt possessing con- 
siderable staying power, but physically weak, with slight muscular 
development, contracted chest and a marked tendency to anaemia, 
which however may be largely due to the innutritious national 
diet of rice, fish, and vegetables. On the other hand the Japanese 
stand intellectually at the head of all Mongolic peoples without 
exception. In this respect they rank with the more advanced 

1 Dr F. H. H. Guillemard in letter to A. H. Keane, Aug. *, 1895. 
8 New York Nation , quoted by the Academy, Sept. 8, 1894. 


European nations, being highly intelligent, versatile, progressive, 
quick-witted, and brave to a degree of heroism unsurpassed by 
any race. The sense of personal honour, so feebly developed 
amongst other Asiatics, became a passion under the mediaeval 
feudal system, and led to astounding acts of devotion and self- 
sacrifice, as well as to deeds of incredible ferocity, of almost daily 
occurrence. With much enterprise and originality is combined 
an imitative faculty surpassing even that of the Chinese, as shown 
by the fact that their first steamer with engines complete was con- 
structed solely from the directions given in a Dutch treatise on 
the subject. These varied mental qualities explain the rapidity 
with which the Japanese, the barriers of exclusion once broken 
down, have taken their place in the comity of the western nations. 
From the mental standpoint the contrast observed between the 
Japanese and their Korean neighbours is all the more remarkable 
since the former are not only physically related to the latter, but 
are also indebted to them for much of their culture. At one time 
Korea was a flourishing centre of the ceramic and other arts, and 
from Chinese or perhaps Manchu materials the natives have 
developed a syllabic alphabet. But for no apparent reason they 
have for centuries been retrogressing, and they now "seem the 
dregs of a race 1 ." Their decay may perhaps be accredited 
to political institutions eminently calculated to yield such re- 

Formerly the expression "Hyperboreans" was collectively 
applied to the Chukchi, Koryaks and other Arctic 
peoples of North-east Siberia, who were supposed bwea 
to form a homogeneous group with the Eskimo 
dwelling under corresponding high latitudes in the New World. 
But if they ever possessed ethnical unity, the Asiatic and American 
branches of this group now stand as widely apart from each other 
as both do from the original Mongolo-Tatar stock. In our 
Family Tree the divergence is shown in the bough to the extreme 
left, where the Eskimo are seen to occupy a branch by them- 
selves, breaking away from the other members of the group soon 
after the common severance from the Mongolo-Tatar connection. 

1 Mrs Bishop, Geograph. Jour. Feb. 1895, p. 162. 



'he Chukchi 

The Chukchi 1 , typical Asiatic " Hyperboreans," are 
scattered in small groups along the shores of the 
Frozen Ocean between the Kolyma River and 
ring Strait, reaching inland as far as the Anadyr basin. They 
re first carefully studied by the members of the Nordenskjold 
;pedition (1878 9), who describe them as tall, lean, with some- 
tat irregular features and fair complexion ; hence they are classed 
de Quatrefages with his " Allophylian Whites," although W. H. 
ill speaks of their coppery tinge 3 . But the statements on this 


m -ir 

*-*^. J ,**^"" 


(In do - Ch inese Type. ) 

d other points are conflicting, owing to the presence in their 
>main of true Eskimo (Chuklukmiut Innuits), with whom some 
servers have confused them. The Chukchi appear, however, to 
ffer altogether from the Eskimo in speech, in the distinctly 
achycephalous shape of the head 8 , and in their light corn- 

1 Properly Tus&i, " Brothers," or "Confederates" (Hooper, Ten Months 
wng the Tents of the Tuski); Nordqvist, however, gives the form Chauchau, 
ural Chauchauate. 

2 Contributions to North American Ethnology, Vol. I. 
8 " Zuin Theil in extremem Maasse" (Virchow). 


plexion. These characters seem to be explained by the statement 
of G. Bove that they came originally from the Amur region 
(Manchuria), and finding the country occupied by the Koryaks 
and the Onkilons, they drove the former south beyond the 
Anadyr, and partly merged with the latter, partly compelled them 
"to cross the Frozen Ocean 1 ." The solution of the Chukchi 
problem, which has been so much discussed, seems therefore to 
be that they are originally a Manchu or Tungus people, who some 
centuries ago settled on the north-east seaboard, where they 
amalgamated with the Onkilon aborigines, that is, with the Ang- 
kali, or so-called "Fishing Chukchis," apparently Koryaks still 
surviving about the Anadyr estuary. The more difficult Eskimo 
problem, thus disengaged from its Asiatic entanglements, will be 
more conveniently dealt with in the next chapter. 

The Tibeto-Indo-Chinese branch, which ramifies to the right 
of the parent stem, presents on the whole far greater The Tibeto 
ethnical unity than the Mongolo-Tatar group rami- Chinese sub- 
fying to the left. On the Asiatic mainland, where 
it occupies almost exclusively the great central Tibetan plateau, 
the southern slopes of the Himalayas, China proper, Indo-China 
and the Malay Peninsula, it may be almost described as homo- 
geneous. Throughout the whole of this vast region, in which are 
concentrated probably one-fourth of mankind, the distinctive 
Mongolic physical type is everywhere in almost exclusive posses- 
sion of the land, the chief exceptions being a few tracts on the 
Southern Himalayas and in Farther India, which are occupied by 
pure or mixed Caucasic peoples. But in the great ocean of 
Mongol humanity flooding south-east Asia for countless ages 
these alien groups may be regarded as une quantite nkgligeable. 

Hence racial problems of such a fundamental and complex 
nature as those of the Ural-Altaic domain do not _ . 


here present themselves. Although physical differ- physical uni- 
ences occur everywhere, they are all confined within rmi y * 
narrow limits. Apart from the few indicated exceptions there 
are no aberrant types, nor even any transitional forms, such as 
those Northern and Western Manchu, Tdrki and Finnish groups, 

1 BoL Soc. Geogr. Ital Dec. 1879, P- 8 3 8 





(Indo-Chinese Tyfe.) 


(Indo - Ch inese Type. ) 




in which Mongolia and Caucasic characters are inseparably 
blended, and which consequently resist all attempts at a scientific 
classification. Tibetans can certainly be at once distinguished 
from Shans, Annamese from Burmans or Chinese, but all alike 
are recognised as not merely Mongoloid but distinctly Mongolic 
peoples, and no question is raised as to their descent from a 
common Mongol ancestry. Thus a yellowish complexion, narrow 
slant eyes, small nose, laterally prominent malar bones, black 
lank hair, and short stature are universally prevalent, and these 
found in combination suffice to constitute a true Mongol type, 
despite discrepancies in minor points. Even the shape of the 
head, which fluctuates within such wide ranges in other divisions 




(Full face.) 

of the Homimdffi, is here somewhat constant, here and there 
mesaticephalous and even sub-dolichocephalous, but mainly show- 
ing a marked tendency towards brachycephaly. 

This general uniformity is well illustrated by the prevailing 
physical and mental characters of the Chinese, the 
most numerous, and one of the most homogeneous 
masses of seething humanity on the globe. Certain 
variations are no doubt presented by the outward traits, such as 
K. 21 

y " 


colour of the skin, ranging from a light lemon and almost white 
shade in the north, to deep brownish hues in the southern pro- 
vinces; the eye usually more or less oblique, but occasionally 
nearly horizontal; the nose normally small and concave, but often 
large and straight especially amongst the upper classes; the 
stature, which though generally under the average, as with all 
Mongolic peoples, often presents surprising discrepancies, as seen 
in the "Chinese dwarfs" and "Chinese giants" from time to time 
exhibited in Europe. But the bony fabric of the Chinese skull, 
so variable elsewhere, is singularly constant both in its resem- 
blances to and differences from the normal Mongol cranium. In 
general it is proportionately longer, and at the same time higher 
than that of any other yellow group, its height slightly exceeding 
its breadth, and the cephalic index falling to about 77*25 (sub- 
dolichocephaly). The face, in complete harmony with the brain- 
cap, is always moderately broad, with very high and prominent 
cheek bones, and jaws developing a slight degree of progna- 
thism 1 . 

Equally constant are the moral qualities, so much so that the 

Mental cha- act i n > f r instance, of a Chinese crowd, may under 

racterofthe given conditions, be always predicted with much 

more confidence than that of any other race at the 

same level of culture. They seem in some respects to be almost 

as incapable of progress as the Negroes themselves, the only 

essential difference being that the arrest of mental development 

comes later in life for the yellow than for the black man. Whether 

this difference is to be explained by a corresponding retardation 

in the closing of the cranial sutures, must remain matter of con- 

1 Dr Hamy, who has made a special study of Chinese craniology, says 
"des machoires projetees en un prognathisme etroit et allonge" (VAnlkro- 
pologie, May, June, 1895, p. -253). This anthropologist here quotes the graphic 
description given by von Baer of the Chinese skull : " Figurez-vous que vous 
ayez un moulage de Fun de ces cranes de Kalmoukes [western Mongols], 
execute" en quelque substance elastique telle que la gutta-percha, et que vous 
comprimiez avec les deux mains chaque cote de la voftte, de fa9on a faire 
monter le front et saillir plus encore le sommet de la voflte et 1'occiput ; 
comprimez plus fortement les arcs zygomatiques, pour qu'ils deviennent plus 
e*troits et que les os jugaux et surtout les maxillaires se profilent en avant, et 
vous aurez le type chinois " (ib.). 


jecture, at least pending further craniological studies in the direction 
indicated on pp. 44 5. 

Meanwhile it may be pointed out' that Chinese culture has 
been stagnant since the early historic period, despite many im- 
pulses from within and without to shake off the chronic state of 
lethargy in which the nation seems content to vegetate. The late 
Terrien de Lacouperie has advanced many arguments to show 
that, before reaching their present homes in the Hoang-ho and 
Yang-tse basins, the primitive Bak tribes had long been in contact 
with the civilised Akkad populations of Babylonia. Hence they 
reached China already a somewhat cultured people, with a know- 
ledge of letters, astronomy, and various industrial arts. In their 
new environment they continued the development of these arts 
up to a certain point, after which, that is, throughout the greater 
part of their historic life, they have mostly remained at a stand- 
still, and even now find the greatest difficulty in assimilating 
Western ideas. This inert mass of semi-civilised savagery offers a 
dead resistance to all outward pressure, even at the peril of the 
national stability more than once overthrown by a few rude Tatar 
hordes. Their religion remains a system of cold moral precepts, 
combined with the old shamanistic superstitions, beneath a veneer 
of Buddhistic ceremonial, ancestry and spirit (demon) worship. 
Their astronomy has scarcely advanced beyond the astrological 
state, while their medical art continues to be a hopeless mixture 
of superstitious practices, absurd nostrums, and a few grains of 
common sense. Excessively courteous amongst themselves, they 
are rude and aggressive towards strangers, with a deep-rooted 
feeling of contempt and even hatred of foreigners and all their 
ways. On the other hand the Chinese, although reckless gamblers 
like all the Indo-Chinese and Malay peoples, are naturally frugal, 
thrifty and parsimonious, which, combined with great staying 
power and capacity for enduring hardships on poor fare, makes 
them formidable competitors with the western nations in the 
labour markets of the world. A characteristic trait is their ex- 
cessive gregariousness, shown in the tendency to crowd together 
in large villages and cities, so that small hamlets and scattered 
farmsteads are scarcely anywhere seen in China. In San Fran- 
cisco 10,000 Chinese are packed together in a space where a 

21 2 


thousand whites would be asphyxiated. This again leads to other 
evils, and especially to a low state of morals, which is one of the 
main objections to the free admittance of Chinese immigrants 
into European colonies. As exaggerated statements continue to 
prevail regarding the population of China, sometimes estimated 
as high as 500 and even 600 millions, it may be remarked that no 
real census has ever been taken. The official estimates of 414 
millions for 1842 and 404 for 1890, are certainly excessive, and 
have been reduced by Herr Kreitner of the Szechenyi expedition 
down to about 150 millions. Other rough calculations give from 
280 to 350 millions, and the population is now generally supposed 
to be about the same as that of British India taken in its widest 
sense, say 275,000,000. 

Further unity is imparted to the Tibeto-Chinese sub-division of 

the Mongol family, by its common isolating form of 
cwneseiin- speech, to which is usually applied the misguiding 
guistic reia- epithet " monosyllabic." It was shown (Ch. IX.) that 

monosyllabism is not the original nor the essential 
condition of these languages ; it is not even a constant character 
in their present state, for imperfect dissyllabic compounds abound 
in Chinese and Siamese, while true polysyllabic compounds and 
derivatives are frequent in the Tibeto-Burmese group. Thus in 
Burmese : kaun^ good ; akaun, goodness ; hlukthan, to ring, from 
hluk, shake and than, sound. In compounds of this type it may 
even happen that neither element is any longer found separate, in 
which case the dissyllable is incapable of decomposition. 

A far more important feature than the length of the words is 
their tonic utterance, the origin and nature of which was neces- 
sarily misunderstood so long as these languages were supposed to 
represent a primitive condition of speech. It is now clear that 
tone gives no support to the theory of a supposed primitive sing- 
song utterance, but that it is a compensating element uncon- 
sciously introduced to distinguish the numerous homophones 
resulting from the ravages of phonetic disintegration. " Thus the 
monosyllable pa will be toned in six or more different ways to 
represent so many original dissyllables, pada, paka, pana, pasa, 
pata..., and some of the Chinese and Shan dialects have, in fact, 
as many as ten or twelve such tones, which unless correctly 


uttered lead at once to the greatest confusion, and to all kinds of 
misunderstandings. Hence these languages are now called iso- 
lating and tonic rather than isolating and monosyllabic*" 

It may perhaps be asked why, this being so, tones in Tibetan 
"eke out scanty inflections, but do not form an Tibetan 
important feature of the language 2 ." The explana- linguistic 

.... 1111 -11 affinities. 

tion is that this language has developed, or possibly 
never lost, numerous grammatical processes which largely dispense 
with the use of tones. Such is regular tense formation by internal 
vowel change, as in hgel, to load; past mkal\ future dgal, impera- 
tive k'ol. Tibetan has this remarkable feature in common with 
the Kottian of the Yenisei basin, Siberia 8 , showing a possible 
original connection between the Ural-Altaic and the Tibeto- 
Chinese linguistic families, or at least a parallel line of develop- 
ment with later divergence towards a flexionless analytical state in 
the Tibeto-Chinese group. Thus might also be explained the 
intricate grammatical forms occurring in some of the idioms on the 
South Himalayan slopes, such as the Vayu 4 , and especially the 
Kiranti of East Nepal, whose complicated verbal system shows 
analogies with the Munda, Sonthal and other Kolarian tongues of 
Lower Bengal 5 . Through the archaic Lepcha dialect of Sikkim 
and Bhiitan, the Bodo (Kachari) and Dhimal of the Terai, the 
Dophla, Miri, Abor and Mishmi of the Eastern Himalayas, and 
the Mikir, Khasi 6 , Garo (closely allied to Kachari) and Naga of 
the South Assam Hills, the Tibetan system passes over to the 

1 A. H. Keane, Population , Races, Langtiages, and Religions of the World \ 
in Chtirch Missionary Intelligencer, October, 1894, p. 723. 

2 Prof. John Avery, Proc. \^th Annual Session, American Philol. Ass. 
July, 1885, p. xvii. 

3 Castren, Yen., Ostiak. und Kott. Sprachlehre. 

4 Specimens by B. H. Hodgson in Bengal As. Jour. XXVI., 1858, p. 372. 

5 Dalton, Ethnology of Bengal, p. 102. "The verb has a remarkable 
development, for, though poor in tense-forms, it has a profusion of forms 
expressive of the relations of subject to object. Participles, too, vary according 
to the tense of the principal verb. Altogether the possible forms of a Kiranti 
verb amount to several hundred " (Avery, ib. p. xviii). 

6 Khasi, however, would appear to be a stock language of peculiar struc- 
ture, having no tones although isolating, and showing scarcely any affinity to 
"the rest of these mountain dialects" (H. Roberts, A Grammar of the lOiassi 
Language, 1892). 


more isolating and toned languages of Indo-China: Burmese, 
Kakhyen, Lushai, Chin, Karen, Mon, Annamese, and the wide- 
spread Shan group of Assam, Upper Burma, South China and 
indo Ocea Siam. But agglutinating forms reappear in Karen, 
nic linguistic while a distinctly polysyllabic group of untoned 
languages, with Oceanic (Malayo-Polynesian) affini- 
ties, occupies a great part of Camboja and surrounding uplands 
(Khmer, Kuy, Charay, Stieng, Cham) 1 . These Oceanic affinities 
have now been traced to the very heart of the continent, and 
T. de Lacouperie confirms B. H. Hodgson's suggestions regarding 
the relations of Gyarung on the Tibeto-Chinese frontier with 
Tagalog, the chief Malay language of the Philippine Archipelago, 
"The Gyarungs were nothing more nor less than one of the 
disjecta membra (now driven away by the pressure of the Chinese 
growth west, south, and also east) of a former nucleus of the native 
population of China, Indonesian in character at the beginning, 
and gradually diverging from their former standard under the 
combined influences of their new surroundings, linguistical and 
others 2 ." 

Here it should be noticed that the term " Indonesian," intro- 
duced by Logan to designate the light-coloured 
nfts!ans. nd " non-Malay inhabitants of the Eastern Archipelago, 
is now used as a convenient collective name for 
all the peoples of Malaysia and Polynesia, who are neither Malays 
nor Papuans, but of Caucasic type'. Such are the Battaks of 
North Sumatra, many of the Bornean Dyaks, most of the Jiiolo 
natives, many of the Philippine Islanders, and the large brown race 
of East Polynesia, that is to say, the Samoans, Maori, Tongans, 
Tahitians, Marquesas Islanders and the Hawaiians, who are 
commonly called "Eastern Polynesians." Dr Hamy, who first 
gave this extension to the term Indonesian, points out that the 
Battaks and other pre-Malay peoples of Malaysia, so closely 
resemble the Eastern Polynesians, that the two groups should be 

1 A. H. Keane, Relations of the Indo-Chinese and Inter- Oceanic Races and 
Languages, passim. 

3 Formosa Notes, 1887, P- 69; see also his Languages of China before the 
Chinese, 129-144, and 225. Here is established "the existence in the east 
of China of dialects of a North Indonesian character." 




regarded as two branches of an original non-Malay stock. 
Although all speak dialects of the common Malayo-Polynesian 
language, the physical type is quite distinct, and rather Cau- 
casic than Mongolic, though betraying a perceptible Papuan 
(or Negrito) strain especially in New Zealand 1 and Mikronesia. 


(Malay Type.} 

The true Indonesians are of tali stature (5 ft. 10 in.), muscular 
frame, rather oval features, high, open forehead, large straight or 
curved nose, large full eyes always horizontal and with no trace of 
the third lid, light brown complexion (cinnamon or ruddy brown), 
long black hair, not lank but often slightly curled or wavy, skull 

1 Yet even in New Zealand Dr O. Finsch met " some full-blood Maoris 
with quite European features, eyes mostly beautiful, full, large, brown to deep 
brown, straight or well curved nose, full beard, and well developed calves, 
as is characteristic of all Polynesians" (Reise in der Siid$ec> 1884, P- 2 5)- 
This is important testimony from an observer who, against his own evidence, 
is inclined to connect both Polynesians and Mikronesians "als Easse" with 
the Malays (p. i). In the same way he confounds the Philippine Negritoes 
with the Melanesians under protest from Herr Virchow, who remarks that 
"em Craniologe wird nicht leicht Hrn. Finsch zustimmen, wenn er die 
Negritos der Philippine!! einfach zu den Melanesiern zieht" (ib. Vorwort, 
p. viii). 


generally brachycephalous like that of the melanochroic Euro- 
pean 1 . 

Thus severed from its unnatural Indonesian connection, the 
Malay problem may have some prospect of a satis- 
factory solution. In some of the early essays at 
classification the Malay race found a place amongst 
the main divisions of mankind (p. 164). Then the very existence 
of a Malay type was questioned by scientific systematists, and 
craniologists especially failed to discover a normal Malay head 
amid the endless discrepancies presented by specimens from the 
Eastern Archipelago. It could scarcely be otherwise when Indo- 
nesian, "Alfuro 8 ," Mikronesian, Polynesian, and true Malay skulls 
were all alike ticketed "Malay" in European collections. Thus 
"the dimensions which Welcher has found in the Malay nations 
are especially surprising.... We find the Maori, with an index [of 
breadth] of 73, still on the verge of dolichocephalism... Marquesas 
74, Tahitians 75, Chatham 76, Sandwich 77, Borneo Dyaks 75, 
Balinese 76, Amboynese 77, skulls from Sumatra 77, Mancassar 
78. To these mesocephali must be added, as brachycephali, the 
Javanese and Buginese with 79, Menadorese 80 and Madurese 
82... Of the 19 gradations of breadth the skulls of the Malay 
family occupy no less than nine. ...According to Barnard Davis the 
Maori (75) are most inclined to dolichocephalism, while the 
Javanese (82) appear still more brachycephalic than the Madurese 

1 Dr E. T. Hamy, Btdl. d. I. Soc. de Geo. XIII. 1877, passim. 

2 This term Alfuro is specially confusing. Whatever its origin, whether 
Portuguese, Arab or local, it never had any ethnical value, being indifferently 
applied by the Malays to all rude non-Muhammadan peoples in the eastern 
parts of Malaysia. So heathen, from heath, originally "rude," "rustic"; 
pagan, paynim, hompaganus, a "villager." Thus C. B. H. von Rosenberg: 
" 26 villages stretch along the coast [of Ceram], of which 5 are inhabited 
by Christians, 3 by Muhammaclans, 15 by Alfuros [neither Christians nor 
Muhammadans], while 13 have a mixed population" (Malayische Archipel, 
Part ii. p. 26). This passage also shows that the " Alfuros " were not necessarily 
the aborigines (as is generally supposed) driven into the interior by the Malay 
invaders, for, here they are found dwelling on the coast peacefully associated 
with their Christian and Moslem neighbours. 

8 Oscar Peschel, Races of Man, p. 55. So also Prof. Sir W. H. Flower: 





(Eastern Indonesian Type.) 


But when the disturbing elements are removed, the true 
Malays are seen to present remarkably uniform 
physical type, characters, and Dr Finsch himself was struck by 
this very uniformity in the subjects from every part 
of Malayland studied by him at Batavia in 1881 '. Thus there is, 
after all, a Malay type, and its characters are such as enable it to 
be at once pronounced distinctly Mongoloid; one might almost 
say Mongolic without reservation, but for the somewhat straight 
nose and large round and generally horizontal or but slightly 
oblique eyes. Yet even here is seen the peculiar Mongol fold of 
the upper lid, "just as with the Chinese," says Finsch (p. 28). 
Other marked Mongol features are very prominent malar bones, a 
dirty yellow or brownish olive colour, very black lank hair, scant 
or no beard, low stature ranging from little over five feet to five 
feet four or five inches, brachy or sub-brachycephalous head 2 . 

Thus is fully justified the Oceanic Mongol group, which in our 
Family Tree is seen to ramify from the Tibeto-Chinese stem east- 
wards to Formosa and south-westwards to Madagascar. The 
Malay affinities of the Formosan aborigines have long been recog- 
nised 8 , and here it will suffice to add that in the island no trace 
has yet been discovered of a Negrito or Papuan element after 
a search of about two hundred years since their 
affmitfe*? 8y existence was first reported by the Dutch. In 
Madagascar, on the contrary, the Malay type even 
of the dominant Hovas has been considerably modified by ad- 

" There is certainly no very great uniformity in the characters of the skulls 
in our collections which are said to belong to Malays" (The Native Races of the 
Pacific Ocean, 1878, p. 41). 

1 "Zunachst war mir bei diesen malayischen Volkern, was allgetneinen 
Typus, Grosse und Hautfarbung anbelangt, die im Allgemeinen herrschende 
Uebereinstimmung auffallend, eine Uebereinstimmung wie ich sie in ahnlicher 
Weise bei SUdsee- Volkern nicht beobachtct hatte. Dieseibe gipfelt vorzugs- 
\veise in dem starken Hervorspringen der Jochbogen, welches Sudsee-Volker 
in weit geringerem Grade, zum Theil kaum zeigen " (ib. p. 27). 

2 Cephalic index of Sundanese 83*9 ; of Javanese 78*2 (J. Deniker and 
L. Laloy, Les Races exotiques, &c., in L'Anlhropologie, 1890, No. 5, p. 543). 

8 Mr Taylor, however, brings some of the tribes from the north, the 
Pepohoans from Liu-Kiu, the Tipuns probably from Japan, &c. (China 
Review, Vol. xvi. No. 3). 


mixture with a Bantu Negroid element from the neighbouring 
mainland. What is specially remarkable in this island is the 
surprising uniformity of speech, a pure Malayo-Polynesian Ian- 
gunge being everywhere current, with but slight dialectic variety, 
amid semi-cultured and rude Malayo-African populations (p. 205). 
It is noteworthy that the relations of Malagasy are not so much 
with the standard Malay, as with some of the more 
remote and more archaic members of the Oceanic Prtynlsian 
family, such as the Kavi, parent of modern Javanese, linguistic reia- 
the Tagalog and Bisayan of the Philippines, the 
Maori, Tahitian and other Polynesian tongues. Thus the numerals 
seven and eight correspond in all these languages (roots pito, valtt)> 
but not in Malay (tujoh, delapari)*. The explanation is that the 
early Oceanic migrations took place in remote times, long before the 
rise and expansion of the Malay nation as now constituted. This 
energetic race of sea-rovers and conquerors had its cradle in the 
Sumatran district of Menangkabau, whence they began to spread 
abroad, apparently not earlier than some eight or nine hundred 
years ago, founding permanent settlements in the Malay Penin- 
sula 2 , around the Bornean seabord and as far east as the 
Moluccas. Through their maritime expeditions and trading rela- 
tions, their simple and harmonious but comparatively modern 
Sumatran dialect became the general medium of intercourse, a sort 
of lingua franca throughout the whole of Malaysia, and even 
in parts of Papuasia and along the Cambojan and Annamese 
coastlands. But there is nothing to show that these later Malays, 
the Orang Malay u* in the stricter sense of the term, ever 

1 "La premiere racine, pitou, dont on ne trouve pas trace en malais, se 
retrouve simultanement en Madagascar, aux Philippines, a Timor, dans la 
Nouvelle Ze"lande et a Taiti" (Aristide Marre, Les Affimtts de la Langue 
Malgache, &c., Leyden, 1884, p. 154). And Melanesia might have been 
added, for " the Melanesian decimal series of numerals is not borrowed from 
the Malay... but is identical with that generally in use in the Indian Archipelago 
and Madagascar" (Codrington, The Melanesian Languages, p. 229). 

2 According to the national records this region was reached by a colony 
direct from Menangkabau in 1238, the foundation of the first settlement, 
Singapore, dating from that year. 

3 This term >J*A* Malayu, which has acquired such a prominent ethnical 
and linguistic position, is of unknown origin, possibly the name of an obscure 


penetrated either into Polynesia, or westwards to Madagascar, at 
least in sufficient numbers to form distinct settlements and 
acquire a dominant influence in those regions. Hence the 
diffusion of the Malayo-Polynesian speech, for which a better 
name would be Indo-Pacific or simply Oceanic, is not due to, 
but long ante-dates, the diffusion of the Sumatran people, from 
whom the inhabitants of the Eastern Archipelago are now named 
and are wrongly supposed to be sprung, or at least to have 
acquired their "Malay" speech. Thus may now be understood 
the otherwise inexplicable phenomenon, that the Malagasy lan- 
guage has on the whole perhaps more intimate relations with those 
of the Philippine Archipelago, of Melanesia, and even of Easter 
Island "within measurable distance" of South America, than with 
the standard Malay of Menangkabau, almost the nearest land in 
Malaysia to Madagascar. All are independent offshoots of the 
common Oceanic speech, which has its roots in Central Asia 
and of which Malay proper is relatively speaking quite a recent 

In the Philippine Islands the conflicting statements of 
observers correspond with the intense ethnical con- 
fusion prevalent amongst the motley populations of 
Philippine the Archipelago. Here "the constant mingling of 

different races from China, Malaya, and parts of 
Melanesia and Polynesia has created a mixture of which the com- 
ponent parts are almost undiscernible V Nevertheless the means 
of introducing some order into this chaotic field seem to be sup- 
plied by the writings of such recent observers as Dr Montano 8 
and Prof. Blumentritt 3 . Apart from the true Negrito aborigines 

Sumatran tribe, which rose to power under some renowned chief in the loth 
or nth century of the new era. The derivation from the Javanese m-layu, 
to flee, cannot be accepted, because the Malays are not "fugitives" but 
everywhere aggressors; nor is it credible that they would accept from 
strangers a term of reproach as their national designation. That this has 
always been their national name is evident from the fact that on reaching the 
mainland in the i3th century they called it Tanah Maldyu, " Malay land," 
whence the present expression Malay Peninsula. 

1 Nature, Oct. 7, 1886. 

8 Voyage aux Philippines, 1885, passim. 

8 Numerous papers in Globus, Vol. 50 and elsewhere. See also Capt. 


dealt with in Chap. XI, Blumentritt distinguishes two separate 
"Malay" invasions, both pre-historic. Montano also recognises 
these two elements, which, however, he more correctly calls 
Indonesian and Malay. The Indonesians, whom he affiliates to 
the "Polynesian Family," were the first to arrive, being followed 
by the Malays and then in f the i6th century by the Spaniards, who 
were themselves followed, perhaps also preceded, by Chinese and 
others. Thus Blumentritt's Malays of the first invasion, whom he 
brings from Borneo, are Montano's Indonesians, who passed 
through the Philippines during their eastward migrations from 
Borneo and other parts of Malaysia. The result of these succes- 
sive movements was that the Negritoes were first driven to the 
recesses of the interior by the Indonesians, with whom they after- 
wards intermingled in various degrees. Then the Indonesians 
were in their turn driven by the M alays from the coastlands and 
open plains, which are consequently now found occupied mainly 
by peoples of true Malay stock. Such are the Tagalas, Bisayas, 
Bicols, Pampangos, Ilocanes and Cagayanes, besides the so-called 
"Moros," that is, the Muhammadan Malays of the Sulu Archi- 
pelago, Palawan and Mindanao. Then with peaceful times fresh 
blends took place, and to previous crossings are now added 
Spaniards and Chinese with Malays, these "quadroons" and 
"octoroons" with Indonesians, and even here and there with 
Negritoes. It has thus become difficult everywhere to distinguish 
between the true Malays and the Indonesians, who are also less 
known, dwelling in the more remote upland districts, often in 
association with the Negritoes, and not always standing at a much 
higher grade of culture. Of these savage Indonesians the tribal 
groups are endless, the more important being the Igorrotes 
studied by Dr A. B. Meyer 1 , the Tinguianes, Guinanes, Apayos, 
Gaddanes, Bagobos, Tagabawas, Samals, and Mandayas. Thus 
the Philippine half-castes may be roughly classed as Negrito- 
Indonesians, Malayo-Indonesians, Malayo-Europeans, and Ma- 
layo-Chinese, the Indonesian element giving here as elsewhere in 
Oceanica the clue to the puzzling ethnical entanglements. 

L. Gatta's summary of Jordana y Morercts investigations, in Boll. Ital. Geo. 
Soc. Feb. 1886, p. 122 et sey. 

1 Eine Weltreise t &c., Leipzig, 1885. 



America peopled from the Eastern Hemisphere during the Stone Ages The 
bronze age of Chimu (Peru) no proof of later intercourse between the 
Old and New Worlds Hence the American aborigines are the direct 
descendants of palaeolithic and neolithic man and their later culture is 
consequently an independent local development But Homo Americanus 
is not autochthonous, but a specialised form of a Mongol prototype 
General Uniformity of the American physical type Texture of the hair ; 
colour of the skin "White" and "Black" aborigines no proof of 
early migrations from Europe or Melanesia Arguments of De Quatrefages 
discussed The Japanese myth exposed The " stranded junk" argument 

Culture of the early Stone Age identical in both hemispheres But 
after that age the arts and industries show continuous divergence in 
America Argument based by Retzius on the two types of American 
crania Contrasts between the present Mongol and American physical 
types Mental Capacity of the American aborigines superior to the 
Negro, on the whole inferior to the Mongol But the Cranial Capacity 
inferior both to Mongol and Negro Striking uniformity of the mental 
characters of the aborigines in North America in South America 
Uniform character of American speech in its general morphology 
Fundamentally distinct from the structure of the languages of the Old 
World Surprising number of American stock languages despite their 
common polysynthetic type Classification of the aborigines must always 
be mainly based on language Family Tree of Homo Americanus 
America probably peopled by two routes From Europe by palaeolithic, 
from Asia by neolithic man Present distribution of the two types 
The Eskimo question Its solution Prof. Mason's theory of the peopling 
of America from Indo-Malaysia Negative Objections to this theory 
Positive Objections True explanation of the coincidences between 
certain usages and mental aspects of the inhabitants of the Old and 
New Worlds Due not to contact or borrowings, but to their common 
psychic constitution Results of the discovery and re-settlement of 
America on the aborigines in Latin America In Anglo-Saxon America 

The Anglo-American type due, not to miscegenation, but to con- 

ELSEWHERE (Chap. X.) general reasons were given for detaching 
the American aborigines from the Mongolic connec- 

America . . 

peopled from tion, and treating them independently, as one of the 
* ur ma * n divisions of the Hominidse. It was also 

during the shown (Chaps. V., VI.) that while a Neolithic age is 

universally accepted for the New World, there are 

also good grounds for accepting a Palaeolithic age for at least the 


southernmost parts (Patagonia, Fuegia) of that region. On the 
other hand there are no records of any migrations between the 
Eastern and Western Hemispheres in pre-Columbian or pre-Norse 
times throughout the historic period, which at all events for Egypt 
and Babylonia goes back some 8000 or 10,000 years from the 
present time (p. 56). Outside those earliest centres of civiliza- 
tion primitive man was at that remote period everywhere at a low 
plane of culture, from which it follows that, if America was 
peopled from the Old World, the occupation took place and was 
practically completed during the two stone ages. 

The general absence of bronze as well as of iron excludes 
those metal periods, while the copper age was in 
the east too short and of too ill-defined a character The bronze 

... T , age of Chimu 

to be here taken into account. Iron was unknown (Peru), no 

except in meteoric form. But bronze implements ^/course* 

in great number and variety have been collected between the 

. , . - f - Old and New 

amid the vast rums of Chimu, a Peruvian city, Worlds. 

capital of an empire overthrown by the Incas 
(Squier, Peru, passim). The occurrence of chumpe, as the alloy is 
locally called, in this district, and nowhere else in the New World, 
is almost equally inexplicable, whether we suppose the metal itself 
to have been prepared on the spot, or only introduced and 
wrought into diverse objects by the local workers in bronze. 
The few bronze objects, little bells and other trinkets, found 
in the Isthmus of Panama and in Mexico, appear to have been 
imported, perhaps from Peru 1 . But for Chimu a real bronze age 
may be claimed. The people were skilled in other arts and their 
earthenware was so beautiful in form and finish that they may be 
called the "Etruscans of the New World." Deposits of tin occur 
both in Mexico and in Bolivia, and .some of the mines appear to 
have been worked in pre-Columbian, times, so that the Chimu 
people may have been expert metallurgists as well as artificers. 
In any case this solitary instance scarcely warrants the assumption 

1 Mr W. H. Holmes suggests that the bronze objects found in some of the 
Chiriqui graves may be post -Columbian, " pointing toward the continuance of 
the ancient epoch of culture into post -Columbian times" {Ancient Art of the 
Province of Chiriqui) in Sixth Annual Report of the Bureau of Ethnology^ 
Washington, 1888, p. 186). 


of direct trading relations between the two hemispheres in the 
bronze age, that is, long before the dawn of Chirnu culture. On 
any later contact history is silent. 

It may thus be inferred that, before the discovery, America 

received no ethnical contributions of any import- 

Hence the _ . , 

aborigines are ance from any quarter after the stone ages, and 

ants^aTao- consequently that the aborigines are mainly the 
lithicand direct descendants of palaeolithic and neolithic 

neolithic man, 

man. If this inference can be established, the 
further inference will follow of itself, that all their arts and institu- 
tions, everything comprised under the general expression "cul- 

ture/ 1 are indigenous, those only being excepted 
culture is a which may be traced to the pre-metal ages. These 
ment dCVelOP " inferences may thus be briefly formulated: Homo 

Americanus branched off from Homo Mongolicus 
in the Stone Ages, and since then has pursued an independent 
local evolution, arrested by the arrival of Homo Caucasicus in late 
historic times. 

It is evident that, owing to the absence of the higher 

apes, the New World cannot be regarded as an 
Americanos is independent centre of evolution for man himself. 
thon a o U u^but a Hence for the American division of the Hominida? 
specialised there is no question of a transition from an anthro- 

formofaMon- ., , . _ _ 

goi prototype. P O1(1 precursor, but only from an already special- 
ised human form. On the other hand the American 
undoubtedly approximates nearest to the Mongol form, and as the 
latter cannot be derived from the former, it follows, as is now gene- 
rally allowed, that the American type has been differentiated from 
a generalised Mongol prototype. Thus is established without any 
lengthy argument, the first assumption of our formula: "Homo 
Americanus branched off from Homo Mongolicus." 

This is also in accordance with physical, geographical and 

General uni- ot ^ r considerations. A strong argument for the 

formityofthe substantial unity of the American race is based by 

*icTi r type Phy " Messrs Flower and Lydekker on the great difficulty 

of forming within the group any natural divisions 

"founded upon physical characters 1 ." Thus the hair is every- 

1 Op. cit. p. 752. 


where black, straight, lank and long, often very long, falling down 
to the waist and even lower. The colour also "varies but little," 
generally presenting different shades of a reddish, olive, or coppery 
brown, whence the expression "Redskins." Although specially 
characteristic of the North American Prairie Indians, this coppery 
tint also prevails in parts o South America, as amongst some of 
the Amazonian tribes, whose "skin is of a coppery or brown. 
colour of various shades 1 ," and especially amongst the natives of 
Guiana, whose colour is described by Mr E. im Thurn as a "very 
red cinnamon" though differing considerably in the different 
tribes 8 . On the elevated plateaux it passes to a more decided 
brown, and in the Brazilian woodlands often to a 
leathery or faint yellowish hue, as amongst the and "black 
Botocudos of the eastern seabord. Both "white" 

and "black" shades are also mentioned, and on early migra- 

tions from 
these terms, which should obviously be taken in a Europe or 

relative sense, some fanciful theories of prehistoric Melanesia - 
and even historic immigrations from Europe and Africa have been 
built. De Quatrefages devotes many pages to the discussion of 
these questions, and although obliged to give up "immigration en 
masse" (op. tit. p. 559), and to "oppose conjecture to conjecture" 
(p. 555), he still believes that Melanesians or Papuans gained a 
footing and maintained themselves on the shores of California, 
because some of the local tribes are spoken of as "black." " The 
faces of the Achomawis," says Mr Powers, "are broad and black, 
and calm and shining with an Ethiopian unctuousness." But we 
had already been warned by La Pdrouse that these Californians 
were in no sense "Negroes,' 1 but obviously of Mongol stock, as 
shown by their lank hair, high cheek-bones and oblique eyes. So 
also with the extinct "black" but lank-haired Charruas of South 
Brazil, and the "white" Antisians (Guarayos, Yuracares) of the 
east Peruvian and Bolivian slopes, these possibly descended from 
some "white Africans'* (Guanches, Biibis) stranded on the Brazil- 
ian coast and penetrating thence across the Amazonian forests to 
the foot of the Cordilleras. Surely it would be simpler to regard 
these "bearded savages" as the result of crossings with European 

1 A. R. Wallace, Travels on the Amazon and Rio Negro* p. 478. 

2 Among the Indians of Guiana^ 1883, p. 189. 

K. 22 











captives, of whom there was no lack during the Indian wars, or 
even as runaway Spaniards adopting the native speech and usages 
as others have done in Bolivia and Yucatan. 

That the reports of white and bearded natives must be 
received with caution, is evident from the current accounts of the 
Mayorunas of the Maranoa (Upper Amazons) and its Ucayali 
and Yavari tributaries, who are also said to have thick beards and 
white skins, and who are supposed to be descended from some of 
the Spanish soldiers left in the district after the murder of Pedro 
de Ursua by Lopez de Aguirre. But it was Aguirre's followers 
who had received the name of Maranones, " People of the 


(Ilispano- American Mixed Type.") 

Marafion," and this word was afterwards confused with Mayoruna, 
the name of a full-blood Indian tribe, who are neither white nor 
bearded 1 . The Spanish Maranones have disappeared, though they 
survived long enough for their European features to be transferred 
by popular report to the Mayoruna aborigines. Similar reports 

1 " L'on ajoutait qu'ils ont encore les traits europeens et la barbe noire tres 
dpaisse. II n'en est rien : loin d'etre fils ou metis d'Espagnols, les Mayorunas 
sont, au contraire, des Indiens de race pure." (Reclus, xvm. p. 550.) 

22 2 


long prevailed regarding the mysterious Guatusos of the Rio Frio, 
Costa Rica, who were said to have fair hair and blue eyes, due to 
contact with the English buccaneers, or with some Spanish 
fugitives. But since they have begun to visit the neighbouring 
markets of San Carlos and San Jose, the Guatusos are found to 
have black hair, dark skins and tyigh cheek bones, like the 
Nicaraguan Chontals to whom they appear to be related 1 . The 
Oyariculets of French Guiana, also reported to be white with blue 
eyes and light beard, are now found to be "like other Indians 3 .'' 

Faint traces of the Norse settlers may perhaps be allowed on 
a t ^ ie nortn " east c ast8 ? but is it not a violent assump- 
nesemyth tion to talk of a "Scandinavian dispersion " over 
exposed. ha if the Nortliern Continent; to bring "Ainu 

whites" to Labrador and Hudson Bay; to build hypotheses on 
the exploded Fusang legend of Chinese Buddhist pilgrims, or to 
take seriously M. Guillemin-Taraire's statement that "the members 
of a Japanese embassy were able to converse right off (a premiere 
vue) with certain natives of Sta Barbara County " [California], if 
not to recognise the rock carvings executed by the neighbouring 
coast tribes, which we are assured are not to be distinguished from 
"objects of a like nature fashioned in Japan " (p. 558). 

Lately Mr O. H. Howarth described before the Anthropological 
Institute some of these "rock inscriptions" which he had seen in 
Sinaloa, West Coast of Mexico, which he also traced to a Japanese 
source, and which "seem likely to furnish an important link in the 
problem of the prehistoric colonization of Central America 4 ." But 
amongst the audience was Mr Daigoro Goh, of the Japanese 
Consulate- General in London, who at once snapped this "link" 
with the remark that "I do not see any resemblance in those 
figures of the inscriptions with the prehistoric characters in Japan 

1 Reclus, XVII. p. 304, English ed. 

8 H. A. Coudreau, Bui. de la Soc. de Glograph. June 15, 1891. 

8 To this source might, for instance, be attributed the high degree of 
dolichocephaly observed amongst the Micmacs of Nova Scotia, although even 
this is regarded by Dr Franz Boas as evidence not of Norse but of Eskimo 
contact. "Archaeological facts tend to indicate that the Eskimo must have 
lived along the coast of New England at one time " (Anthropology of the North 
American Indians > p. 45). 

4 Jour. Anthrop. Inst. February, 1894, p. 226. 


known as Hibunci or sun letters ; nor do I find any similarity in 
the Ainu writing. Moreover, I should like to remind the lecturer, 
who said that one of the inscriptions has the crest of the Prince of 
Satsuma, that that identification will not give any weight on [to] 
his assertion, since the former must be a thing of several thousand 
years old, whilst the latter has had only a seven or eight centuries 
existence" (p. 231). Unfortunately Japanese scholars have only 
lately begun to take part in discussions of this nature, so that the 
numerous other links like those of Sinaloa and California which 
abound in uncritical ethnographical writings still remain to be 

Meanwhile it may be pointed out that all these fancied early 
historic relations of the natives with the Asiatic The * strand- 
peoples are not only unsupported by any direct edjunk" 
evidence, but are otherwise involved in tremendous * me 
difficulties. Because a stray Chinese or Japanese junk may have 
occasionally been stranded on the western seabord since the 
discovery, it is argued tliat similar waifs may have arrived in 
remoter times, and given rise to the local cultures. But there 
were no craft capable of traversing the Pacific Ocean in the 
neolithic age, when America was already strewn with monuments 
from the Mississippi basin to the Argentine pampas. At that 
remote epoch, without going still further back to the "discredited" 
palaeolithic man, there were neither specialised Japanese, who 
according to the national traditions reached their present homes 
less than 3000 years ago, nor specialised Chinese, who according 
to T. de Lacouperie migrated from Western Asia to the Hoang-ho 
valley since the rise of Akkad culture in Babylonia. And if any 
of these historical peoples ever arrived in sufficient numbers to 
build up a civilization in the New World, the Asiatic origin of such 
a civilization would be self-evident, and not the subject of heated 
debate between different schools of learned archaeologists. Man 
cannot separate himself from his immediate associations, and the 
eastern founders of such communities must necessarily have 
brought with them their arts, their speech and written records, 
their domestic animals, their more useful cereals and other plants, 
without which they must have themselves speedily perished or 
been absorbed in the surrounding native populations. But no 


trace of these things was found in any part of the New World on 
the discovery. There was neither the rice of the Malays and 
Japanese, nor the tea of the Chinese, nor yet the wheat, barley, 
oats or rye of the western nations, nor the horse, camel, ox, sheep, 
goat, pig or poultry of the Eastern Hemisphere, nor the iron now 
proved to have been known to the ancient Egyptians and 
Assyrians; lastly, not a single written document nor an echo of 
the speech of any of the Asiatic, African or European peoples. 
All was of indigenous growth, maize, potato, llama, mounds, casas 
grandes, Toltec, Nahua, Maya, Peruvian and Aymara monuments 
and languages, man himself, at all events since the stone ages. "To 
say that the Americans are derived from the Chinese, the Japanese, 
the Malays or the Polynesians, is highly unscientific. Theo- 
retically it is probable that the language, the physique, the social 
and religious culture, and the geographical distribution of all these 
peoples, have undergone radical changes since that early time, and 
that since their present stages or any approximation to them have 
been attained, migration to America has not been in progress 1 ." 

But it may be asked why these migrations should be arrested 
at so early a date, and not continued into later times, when man 
might be supposed better equipped for such peaceful or hostile 
movements ? Two answers may be given to this question, which 
is often raised, but usually allowed to go unchallenged. In the 
first place it might suffice to observe that there is no evidence, 
where abundant evidence should be forthcoming, that any later 
migratory movements did take place between the Eastern and 
Western Hemispheres. The proofs relied upon by the advocates 
of Asiatic or European influences are invariably found, when 
critically examined, to possess no weight, while many must be set 
aside as palpable frauds. Such are the stone carvings from Mount 
Pisgah, North Carolina, some specimens of which were brought to 
Europe by the late Mr Mann S. Valentine of Richmond in 1882, 
and exhibited at the London and Berlin Anthropological Societies 2 , 
About the good faith of Mr Valentine himself there never could be 
any doubt. But it has since been ascertained that "these articles 

1 De Nadaillac, Prehistoric America, English ed. p. 523. 

2 A. H. Keane, On North Carolina Stone Carvings, Jour. Anthrop. 
Institute, June 1882. 


were made from the soapstone found in that region by some 
persons who had learned how to give them the appearance of 
age. ...As a proof of the correctness of his statement Mr Emmert 
[of Washington] had the same parties who stated they had made 
some exhibits for Mr Valentine, make quite a number of similar 
articles for the Bureau V A similar object-lesson is afforded by the 
famous " Lenape Stone," to which Mr Mercer has devoted a special 
monograph, without convincing the scientific public that it is any- 
thing more than a clumsy copy of a genuine mammoth carving 
found in the cave of La Madeleine, Perigord, in 1864*. The 
monuments of North America and the associated objects were 
never observed with more intelligent eyes than those of the 
traveller, Bartram, whose conclusion was that "none of them 
discover the least signs of the arts, sciences or architecture of the 
Europeans, or other inhabitants of the Old World ; yet evidently 
betray every mark of the most distant antiquity 3 ." 

In the second place, although later and more civilised peoples 
were undoubtedly better equipped for spreading abroad than were 
those of the Stone Ages, they lived under different conditions, by 
which the difficulties of migratory movements were immeasurably 
increased, and in some regions rendered practically impossible. 
When man first became specialised, he ranged, like the surround- 
ing faunas and floras, slowly but steadily over the still unoccupied 
spaces. He drifted, so to say, unconsciously hither and thither, 
impelled or attracted now in one direction, now in another, by 
various causes, such as overpeopling, changed climatic relations, 
greater or less abundance of food and facilities for obtaining it. 
He thus gradually filled all the inhabitable parts of the earth, 

1 Cyrus Thomas, Twelfth Annual Report of the Bureau of Ethnology -, 
Washington, 1894, p. 347. Here also may be seen an exposure of the 
Davenport and other inscribed tablets written on some eclectic system in 
various Old World scripts, and from time to time extracted from the mounds 
where they had been deposited by the " authors" for the purpose of mystifying 
the credulous archaeologists of North America. " A consideration of all the 
facts leads us, inevitably, to the conclusion that these relics are frauds, that is 
they are modem productions made to deceive " (pp. 642 3). 

2 H. C. Mercer, The Lenape Stone> or the Indian and the Mammoth. 
New York, 1885. 

* Travels i 1791, p. 522. 


following the lines of least resistance, and like the waters that 
seek their level, overflowing into all the empty spaces. But when 
these spaces were themselves flooded by the tide of humanity, 
there necessarily ensued a period of rest, followed at intervals by 
the ebb and flow of fresh currents setting in all directions. To 
the first movement correspond the two Stone Ages, when in fact 
the whole world, America included, was peopled to its utmost 
inhabitable limits (Chaps. V. VI.) 1 . To the later movements of 
ebb and flow correspond prehistoric and historic times, when, the 
empty spaces being already Occupied, every advance involved a 
conflict, in which those perished who were least fitted for the 
struggle. But before the development of navigation insular re- 
gions, such as America, could scarcely be approached at all in 
sufficient numbers to overcome the dead resistance of the more or 
less dense populations in possession of the favourable districts. 
Even the Norsemen failed to effect a permanent footing, and it 
must now be obvious that small bands arriving at intervals in 
praus or junks from the Asiatic seabord could produce no appre- 
ciable impression either socially or ethnically, but must have been 
successively absorbed by the surrounding aborigines. The few 
hyperboreans that may have crossed over by Bering Strait in later 
times could have no influence of any kind beyond the " Eskimo 
fringe," while the crews of any European vessels stranded on the 
inhospitable Brazilian coastlands could do little except supply a 

1 " You know that before there was a beast of burden humanity had found 
its way over the earth on foot, and that in the simplest craft, without compass 
and with only Nature's pilots, every water had been traversed and every 
habitable island in all the seas had been discovered and settled. It is a long 
journey from the supposed cradle land of our species to Tierra del Fuego ; but 
it had been successfully accomplished in prehistoric times" (O. T. Mason, 
Similarities in Culture, in The American Anthropologist, vm. April, 1895, 
p. 102). With regard to the islands, however, it may be pointed out that 
many, such as those of the Eastern Archipelago, were certainly connected by 
continuous land with the adjacent continents in comparatively recent geological 
times. In the Pacific Ocean, also, some, such as New Zealand, occupied far 
wider areas than at present, thereby proportionately diminishing the distances 
to be navigated. Groups and solitary islands far removed from all land the 
Mascarenhas, the Galapagos, St Helena, Ascension, &c.had never been 
reached by primitive man, and were found uninhabited when discovered in 
recent times. 


meal for the ferocious Tamoyo and Botocudo cannibals. Thus 
there are positive as well as negative reasons for believing that 
after the Stone Ages the American aborigines remained secluded in 
their insular domain without any serious contact with the peoples 
of the Old World prior to the discovery. 

Although Virchow's statement 1 may be true that the most 
practised archaeologist will fail to detect any material 
difference between the stone implements of the two Culture of 

..... . the early Stone 

hemispheres, this merely implies that the arts of Age identical 
Palaeolithic and Neolithic man were pretty much spheres!^""" 
alike everywhere, and that, as here maintained, the 
peopling of America dates from and ceases with the Stone Ages. 
But divergencies already appear in neolithic times, and the rude 
ornamentation of the potsherds found in the New England shell- 
mounds shows little resemblance to that of the oldest European 
pottery 2 . The stone implements are identical; the beginnings of 
decorative art already differ. The inference is obvious America 
owes nothing to the Old World after the Stone Ages, since when it 
has pursued an entirely independent ethnical and social evolution, 
undisturbed by, and unconscious of, the occasional arrival of a 
stray Japanese junk, Malay prau, or solitary Buddhist wanderer. 

Hence despite certain apparent coincidences and analogies 
due to the fundamental unity and common psychic 
nature of man the local arts, and social and reli- 

gious institutions continue to diverge in proportion arts and in - 

,,., , /., , dustries show 

as tney reach higher planes of culture. " That the continuous 
Toltec builders of the low truncated Mexican America^ in 
pyramids were a different people from the pyramid 
builders of the Nile Valley, and that the mummies of the Ancon 
necropolis and other parts of Peru were of a different stock from 
the Egyptian mummies is sufficiently evident from the texture of 
the hair alone. The hair of the old cultured races of America was 
the same as that of all the later American races, uniformly lank, 
because cylindrical in section. The hair of the old Egyptians, like 

1 Anthropologie Amerikd's in Verhandlungm der Gesellschaft fur Anthro- 
pologie, 1877, pp. i44-5<5. 

2 Peabody Museum Report^ 1871. The types and processes were already 
widely diffused, as far south as Florida and west to Illinois and Missouri. 


that of the modern Fellahfn, is on the contrary uniformly wavy, 
because more or less oval in section. The religions again, of the Red 
Man, we are told by Carl Schultz-Sellack, Oscar Loew, and other 
good observers, are * essentially astrological, based on star, sun 
and moon worship,' with which was often associated an intricate 
method of measuring time built on p series of twenty constella- 
tions 1 .' The sun, says Loew, 'is the god of most Indian tribes. 
He diffuses warmth and nourishment for us and our animals ; why 
shall we not worship him? observed to me on one occasion 
Masayamtiba, a Moqui Indian 1 (ib. p. 265). This Masayamtiba 
was a better philosopher than those ethnologists who seek for the 
origin of such a simple cult in the remote corners of the globe, 
rather than in the beneficial influence of the heavenly bodies which 
shine alike for all mankind. The four great gods of the Mayas, 
the 'props of the heavens, 1 answered to the four great Mexican 
gods of the four quarters of the compass, all being associated with 
the four elements of wind, water, fire and earth. But to what does 
either system answer in the polytheistic creeds of the Hindus, 
Assyrians, Babylonians, or other nations of antiquity? There is 
something similar in the Neo-Buddhistic teachings ; but Buddhism, 
even of the oldest type, is much too recent to explain anything in 
the religious worlds of Mexico or Yucatan. Waitz* well observes 
that a common belief in a universal flood, or in the periodical 
destruction of the world, whether by fire, water, storms or earth- 
quakes, and analogous or parallel lines of thought, afford no proof 
whatever in favour of affinity 8 ." 

Such affinities with what de Nadaillac calls the "full-fledged 
races" of the Eastern Hemisphere have been sought 
by anthropologists in the shape of the skull. 
ziusonthetwo Andreas Retzius amongst others grouped all the 
American ' American aborigines in two great divisions : i. The 
western highlanders, occupying the Rocky Mount- 
ains and the Andes with the intervening Pacific seabord; 2. All 
the rest, mainly lowlanders, from the western uplands to the shores 

1 Zeitschr.fur Ethnologie, 1879, p. 209. 

2 Anthropology, p. 255. 

8 A. H. Keane, American Indians in Encyc. Britannica, ninth ed. p. 823. ' 


of the Atlantic. The highlanders, assumed to be all round- 
headed, he classed with the brachycephalous Mongols and 
Malays; the lowlanders, assumed to be long-headed, he traced to 
possible Berber and Guanche migrations from North- West Africa 
and the Canary Islands. We have seen (Chap. XI.) that Europe and 
North America were probably connected by continuous land in 
miocene if not later times, and Mme. Marie Pavlovna has recently 
shown that the close resemblance between the Eurasian and 
American mastodons adds much force to the hypothesis of a 
connection between the two continents during the Tertiary period 1 . 
This will account for the peopling of the New World in pleistocene 
times, but it will give no support to the later movements of 
migration implied in the Swedish anthropologist's generalisation, 
the postulated tertiary continent having vanished in the low 
latitude of the Canaries if it ever extended so far south long 
before the arrival of the cultured Guanches in the Archipelago. 

It is also to be noticed that South America was already 
occupied by both long- and round-headed races (Lagoa Santa, 
Pampas, p. 99) in the first Stone Age. Since then America, like 
other parts of the globe, has been the scene of constant ethnical 
movements, shiftings, dislocations and dispersions, so that it would 
be surprising to find the two elements now disposed in the sym- 
metrical order assumed by Retzius. The sharp distinction drawn 
between brachycephalous highlanders and dolichocephalous low- 
landers has in fact no substantial basis, and a closer study of the 
aborigines, after making every allowance for the practice of arti- 
ficial cranial deformation which is wide-spread in some regions, has 
placed beyond doubt the intermingling of cranial types almost to 
as great an extent in America as in Malaysia itself. Thus Prof. 
Kollmann* finds for the northern Continent, excluding Mexico, 
1575 per cent, dolicho; 40-26 meso; 25-81 brachy; 11*96 hyper- 
brachy; and 4*48 ultra-brachy through deformation, without any 
marked relation to geographical areas. According to de Quatre- 
fages and Harny 3 the Algonquians are sub-bra chy in the north, 

1 Bull, de la Soc. des Naturalistes de Moscou, 1894, No. 2. 

2 Die Autochtontn Amerika's in Zeitschrift fur Ethnologic^ 1883, ? ' 
et seq. 

3 Crania Ethnifa,. p. 469 et seq* 




nearly dolicho in the south, and intermediate in the west. Both 
the Mexicans and Peruvians, who ought to be brachy, are sub- 
dolicho (78-1 and 787, Broca), and MM. Deniker and Laloy find 
that the Aztecs are also dolicho or intermediate, "never brachy- 
cephalous 1 ." All the Eskimo irrespective of locality are highly 
dolicho, but increasing from the wesf (75*3) and centre (75*1) to 
the east (7i*3) 2 . Nine Omahas (Dakotan lowlanders) measured 
by M. Manouvrier gave a mean of 83-8, and the Dakotans of 
Col. Cody's troop measured by MM. Deniker and Laloy (ib. p. 
541) a mean of 8o - 66; yet all these ought according to the theory 


(Mil i sea Type.} 

to be dolicho. In South America, similar contradictory results 
have been obtained, and here the lowland Charruas (brachy) 
change place with the highland Muiscas (dolicho). Recently 
Dr Ten Kate 3 found 119 Araucanian skulls of the La Plata 
Museum (near Buenos Ayres) to be distinctly brachy, while the 
neighbouring Fuegians are classed by de Quatrefages as dolicho. 

1 L* Anthropology* September October, 1890, p. 542. 

2 Dr Barnard Davis, Thesaurus Craniorum ; Rink ; Boas. 

8 Quoted by Dr Brinton, Science, new series, Feb. i, 1895, p. 128. 




Excluding deformations, Topinard 1 gives for North America a 
mean of 79*25 and for South America 79*16, both mesaticephalous, 
and consequently implying mixture everywhere. Subjoined is a 
table of results for some of the chief peoples of both continents : 




Calaveras (fossil) 


Pampas (fossil) 



















Sumadouro (fossil) 


















between the 
present Mon- 
gol and Ameri- 
can physical 

Other marked physical characters, showing divergence from the 
present Mongol type, are : i. The well-developed 
superciliary ridge and retreating forehead; 2. Large 
high-bridged nose, often aquiline or showing in pro- 
file the typical busque form, that is, two lines meet- 
ing on the bridge at an obtuse angle, and generally 
leptorrhine; a very general feature showing in all respects the 
greatest possible difference from the Mongol with a close approxi- 
mation to the Caucasic type; 3. Small sunken eye, round and 
generally horizontal, and without the Mongol fold except in the 
aberrant Eskimo group, although here and there "the eyelids 
exhibit all the varieties observed in Asia, being sometimes con- 
tracted and oblique 2 ." 4. Stature distinctly above the average, 

1 Anthropology i p. 240. 

2 Topinard, Anthropology^ p. 479. 

Yet MM. Deniker and Laloy failed to 


with a mean of about 5 ft. 8 in. or 5 ft. 9 in., rising to 6 ft. in 
the Patagonians, and falling to 5 ft. in the Fuegians and some 
Eskimo at the two extremities of the continent. All the Redskins 
are tall, and a large number (517, mostly Iroquois) measured by 
Dr Gould gave a range of from 5 ft. 3 in. to 6 ft. 3 in., with a 
mean of about 5 ft. 9 in. l 

In their general physique, even more perhaps than in these 
details, the average American Indians present the sharpest con- 
trast to the Asiatic Mongol. The physical appearance of Attila's 
Finno-Tatar hordes (Huns and others) caused the deepest aversion 
in Procopius and other western writers, whose vivid descriptions 
were remembered when the descendants of the same fierce 
nomads again burst into Europe some centuries later. But the 
American Redskin often rises to an ideal standard of manly 
beauty, not merely in the glowing pages of Fenimore Cooper, but 
in such personalities as the Apache chief, Geronimo, described by 
General Sherman as "more than six feet in height, straight as an 
arrow, superb in his physique, with long black hair hanging 
profusely about his shoulders and adorned with eagle's feathers 
a splendid specimen of his fast-vanishing race 2 ." Such a picture 
was never yet inspired by the presence of any full-blood Kalmuk 
chief or Mongol khan 8 . 

detect these traits in the group examined by them : " Dans aucun cas nous 
n'avons observe d'yeux a forme mongoloide " (he. cit. p. 543). This Mongol 
eye, however, has been noticed in the women and children of the Omahas 
(de Quatrefages, n. p. 551). 

1 Investigation in the military and anthropological statistics of American 
soldiers, 1869, passim. 

2 With this may be compared Yvon of Narbonne's vivid though no doubt 
overdrawn description of the Tatar hordes contained in a letter to Giraldus, 
Archbishop of Bordeaux (1243), and preserved in Matthew Paris: Habent 
autem pectora dura et robusta, facie* macras et pallidas, scapulas rigidas et 
erectas, nasos distortos et breves, menta prominentia et acuta, superiorem man- 
dibulam humilem et profundam, denies longos et raros, palpebras a crinibus 
usque ad nasum protensas, oculos inconstantes et nigros, aspectus obliquos et 
torvos, extremitates ossosas et nervosas, crura quoque grossa, sed tibias 
breviores, &c. (Chron. iv. R. Luard's ed. 1877). 

3 Struck by these contrasts some anthropologists have gone so far as to 
deny any physical connection at all between the Mongol and the American 
divisions. Dr Brinton amongst others claims to have disproved what he calls 


Despite this physical superiority, the American aborigines are 
generally held to be intellectually inferior to their 
remote Mongol kindred, but greatly superior to city C of the* 1 **" 

Homo ^Ethiopicus. The latter assumption needs no 
proof, being established beyond all question by the 
most cursory glance at the sjpcial evolution of the Black and Red 
races since the Stone Ages. Some groups both in Africa and 
America Negritos, Bushmen, Botocudos, Yahgans still stand 
at the lowest level. But while the New World has been strewn 
with prehistoric remains from the northern prairies to the southern 
pampas, the Negro domain has nothing to show more permanent 
than the wooden "Assembly Halls" of Mangbattuland. At the 
time of the discovery the American Indians pre- 
sented every grade of social progress from the utter 
savagery of the Brazilian forest populations, and 
the partly agricultural state of the hunting tribes of the northern 
steppes 1 , to the more or less civilised Pueblo Indians and 
inhabitants of the Anahuac, Yucatan, Colombian and Andean 
plateaux, merged together in great nationalities, and dwelling 
in flourishing cities, whose wealth and splendour excited the 
astonishment while stimulating the greed and rapacity of the 
conquistadores. When it is remembered that some of these 
cultures were the outcome of slow and independent growth on 
bleak or arid tablelands, developed without the aid of iron or of 
any more useful domestic animal than the feeble Peruvian llama, 
it may be doubted whether the verdict which places the more 
favoured Mongoloid Asiatics above the American aborigines is 

11 the alleged Mongoloid resemblances of the American race," and is severe 
on Dr Ten Kate for still upholding "the Mongoloid Theory" (On various 
supposed relations between the American and Asian Races* 1893, p. 145 and 
elsewhere). But the resemblances are patent, perhaps even more so in the 
southern than in the northern Continent. 

1 Speaking of the northern Continent, Mr J. W. Powell says : " The 
practice of agriculture was chiefly limited to the region south of the St Law- 
rence and east of the Mississippi. In this region it was far more general 
and its results were far more important than is commonly supposed... though 
unquestionably the degree of reliance placed upon it as a means of support 
differed much with different tribes and localities " (Indian Linguistic Families, 
&c., 1891, p. 41). 


entirely justified. It may be allowed that there is nothing in 
Mexico, Yucatan, or Peru comparable to the stupendous temples 
of Boro-Bodor and Angkor- Vat in Malaysia and Indo-China ; but 
these structures were planned by Hindu, that is, Caucasic, 
missionaries, and cannot be credited to the genius of the surround- 
ing Mongoloid peoples. In respect of letters and literature, 
however, the superiority of the Mongol intellect cannot be 
questioned. Neither the Aztec nor the Maya pictorial or ideo- 
graphic writings, nor the Peruvian quipos, nor yet 
suc ^ i nc herent compositions as those of the 
Quichd Popolvuh, written after the Conquest, are in 
any way comparable to the libraries of moral, religious, historical 
and even poetic works produced in China, Japan, Tibet and other 
Mongol lands during the last 1500 or 2000 years. 

Measured by this test the mental capacity of the American 

aborigines is as inferior to that of the yellow race 

But cranial as i s their cranial capacity, as determined by Mor- 

capacityin- r /> J 

fenorbothto ton Mongol average 1421 c.c., American average 
Negro? and I2 34 c - c - But, as already shown (p. 43), measure- 
ments of cranial capacity yield strangely contra- 
dictory results, and this is specially the case as regards those of 
native American subjects. Thus the average here quoted is the 
same as that given for the Oceanic Negroes (Papuans), whereas 
that of the African Negroes rises to 1364 c.c., which is higher even 
than the Mexican (1339 c.c.), and very much higher than the 
Peruvian, which is the same as the Papuan (1234 c.c.) 1 . Yet no 
one would pretend to place the Congo natives intellectually on a 
level with, much less above, the civilised nation whose empire 
under the Incas extended from Ecuador to Chili, and from the 
Pacific coast across Bolivia inland to Argentina. 

Such profound physical and mental contrasts as are here 
indicated between the American and the Mongol divisions can 
be explained only by divergence of the American branch from the 
remotest times, and its subsequent independent evolution in a 
practically isolated environment. Thus is established the second 
part of our formula on physical and mental grounds. The same 


inference is arrived at by a closer study of the mental characters 
themselves, and especially the temperament and speech, of Homo 
Americanus, both of which present a surprising degree of uniform- 
ity, and a no less surprising difference from the corresponding 
characters of Homo Mongolicus. 

All observers are unanimous in attributing to the American 
aborigines a mental disposition marked by slow- 
ness of excitability, and power of passive resist- 
ance, combined with an impassive exterior, a mental charac- 
capability of endurance and self-control, and a 
general wariness carried to a higher pitch than in any other 
division of the human family. This picture is completed by an air 
of sadness or gloom, observed especially in the more cultured 
groups the Aztecs, Quechuas, Aymaras, etc. and obviously 
attributable to the consciousness of a lost past and hopeless 
future. The heroes of romance are grave, solemn, cautious, 
reserved, observant beneath an outward show of indifference, 
steeled by long inheritance and discipline to inflict, or, if van- 
quished, to endure, the most terrible of fates, death by slow 
and excruciating torture. 

The phlegmatic temper of the Greenland Eskimo was already 
noticed in the last century by Pastor Egede, who tells us that 
"they seldom give way to passion, or are much affected by any- 
thing/' but "come and go, meet and pass one another, without 
interchanging any signs of recognition 1 ." "A grave 
demeanour, slow action, and pulse less rapid than Amerfca? 1 * 
the inhabitants of the Old World 9 ," are the distinc- 
tive attributes assigned by M. Reclus to the aborigines generally, 
while wariness is declared to be "the dominant quality of the 
Indian hunter. He searches space with a scrutinising glance, 
notices the trace of footsteps on the ground, studies the crumpled 
leaf and twisted branch, lends his ear to distant sounds, ceaselessly 
questions surrounding nature, and in it reads the brewing storm. 
His mind is ever on the watch, his imagination ever rich in 
stratagem, his patience still unflagging. He can glide stealthily 
through the foliage, drift with the floating log, creep round to 

1 Description of Greenland, p. 122. 

2 Universal Geography, English ed. XV. p. 48. 

K. 23 


leeward of the game, catch the scent and, undetected, crawl 
through the grass to take him unawares. With the enemy, or 
even with the stranger, who may still be a foe, as the pale-face is 
for the most part, he is still the crafty hunter. He keeps on 
his guard, and hides his feelings under an impassive counte- 
nance; seeming neither to hear nor to understand, he sees all 
and remembers what may be needed to ward off or anticipate 
attack. Should he fall into the hands of a stronger or more 
cunning adversary, his mind is already made up. He feels 
that it is due to himself, due to his tribe, still to maintain his 
haughty bearing, still to defy his captors. The early writers tell 
us how, chained to the stake, he urged the women and children 
to tear his flesh, to sever his limbs, to burn him at a slow fire, 
and how, feeling the approach of death, he intoned his war-song, 
so that his last breath might still be a death-rattle of scorn and 
pride 1 ." Such scenes, unparalleled elsewhere, are no fancy pictures ; 
they have been actually witnessed by white men even in the 
present century 2 . Equal endurance is displayed by young and 
old under their fearful ordeals and self-inflicted tortures, such as 
those of which George Catlin was a spectator during his resi- 
dence amongst the now extinct Mandans of the Missouri valley 3 . 
The scenes described by that observer are of such a harrowing 
nature as almost to pass the bounds of credibility, and indeed 
some of the trials of endurance have been questioned or declared 
impossible on physiological grounds alone. Nevertheless Catlin's 
veracity, impugned by Schoolcraft and others, has been confirmed 
by independent evidence. A few of the details must certainly be 
rejected as absolutely incredible; but these are given un the 
hearsay report of "several traders" (p. 368). 

1 Op. dt. xvii. p. 30. 

2 See J. P. Dunn, Junr., Massacre of the Mountains, &c., 1886, p. 513. 

3 North American Indians, I. p. 170 et scq. Catlin's account of the 
appalling cruelties witnessed by him at the Manclan annual ceremonies is 
reprinted with the original illustrations in the Smithsonian Report for 1885, 
Part ii. p. 356 et seq. Here is also published a summary of the controversy 
to which his statements gave rise, together with confirmatory evidence and 
remarks by the editor, who accepts "the correctness of his descriptions/ 1 
and declares him to have been " an honest observer and truthful chronicler " 
(P- 374)- 


Similar mental traits characterise the Central and South Ame- 
rican Indians, such as the Caribs, who " have a 
gravity of manner and a certain look of sadness America, 
which is observable among most of the primitive 
inhabitants of the New World 1 ." 

On the banks of the Paraguay Mr E. F. Knight witnessed a 


(True Car II) Type.) 

scene which reads like an extract from The Last of the Mohicans : 
" We saw four Indians come stealthily down to the bank, armed 
with long lances. Then, lying down among the reeds, they gazed 
silently into the water till they saw some big fish pass by, when, 
with wonderful skill, they speared them one after the other, and 
threw them on the bank. Next, they lit a fire, roasted the fish 
they had caught, and devoured them. This done, they picked 
up their weapons, and crept back into the woods as noiselessly 
and stealthily as they had come. The whole time some three 
hours that they were on the river-bank, not one of these men 
spoke a word 2 ." 

1 A. von Humboldt, Personal Narrative, ill. p. 74. 

2 Cruise of the Falcon, 1884, Vol. II. p. 27. 




. [CHAP. 

In the " strange and painful whip dance " described by Mr 
Everard im Thurn the Arawaks " lash each other until their calves 
are striped with weals and the blood flows freely 1 ." The same 
observer tells us that the Guiana Indians before shooting rapids 
and on other occasions propitiate the local spirits by rubbing red 
pepper in their eyes, and that tfye older people "inflict this 
self-torture with the utmost stoicism." The extreme pain of 
the operation, " which is never omitted," is shown by the fact 


(Arawak Type.} 

that in the children and even young men it causes sobbing, an 
otherwise rare sight amongst the impassive and unemotional 
natives 2 . But the power of endurance, and of uncomplaining 
submission to the direst calamities, shown by the Guarani Indians 
of Paraguay during the war of extermination waged by Lopez 
against Brazil and her allies was never surpassed, scarcely ever 
equalled, by any other nation. " On the battle-fields the allies 
found little but dead bodies ; nor all of these, for many, fighting 
lassoed round the waist by cords attached to the saddle-bow, were 

1 Among the Indians of Guiana, 1883, p. 326. 

2 ib. pp. 368-69. 


borne dead or dying from the field by their mounts. Prisoners 
tore the bandages from their wounds.... The manhood of the 
nation almost entirely disappeared by war, famine and cholera. 
None survived except the infirm, the women and children 1 ." In a 
'word, watchful, reserved, impassive, enduring, gloomy, sullen, are 
the epithets most frequently applied by travellers to the natives of 
South, as well as of North, America, and few will dissent from the 
contrast drawn by Darwin " between the taciturn, even morose 
aborigines of South America and the light-hearted talkative 
negroes 8 ." 

Almost equal uniformity pervades the general morphology of 
American speech, although recent research tends 

. Uniform 

to show that what Dean Byrne calls its " megasyn- character of 

thetic or massive character 3 '* is not by any means 

so universal as is commonly supposed. Never- 

theless this character, the nature of which has 

already been explained (Chap. IX.), is conspicuous in Eskimo, 

1 Reclus, Vol. xix. English ed. p. 295. 

8 Descent of Man, I. p. 216. 

3 General Principles of the Structure of Language, 1885, I. p. 136. In 
this learned work an attempt is made to establish a correlation between the 
mental qualities of all races and the peculiar character of their respective 
languages. The theory is supported by a vast amount of research and acute 
reasoning, and the author's conclusions may perhaps be said to agree better 
with the relations prevailing in the New World than in the eastern hemisphere. 
The general principle is laid down that " slowness and persistence of mental 
action must tend to impede the movements of thought which are involved 
in language, and to make its acts larger so as to embrace a wider object" 
(i. p. 22) ; and it is claimed that the theory is proved for America by the 
massive character of its speech, corresponding to the slow mental action of 
the aborigines. Despite its inductive treatment, the subject belongs, and 
must long belong, to the region of metaphysical linguistics. Its general 
conclusions seem to be vitiated, amongst other considerations, by the pheno- 
menon of speech shifting from one race to another (p. 202) without such a 
corresponding mental transformation as would be necessitated by the hypo- 
thesis. The English-speaking Irish Kelts have not acquired a Teutonic 
habit of thought, nor has the English language spoken by them made any 
appreciable approximation to the general structure of Keltic speech. It 
would, on the other hand, be difficult to show that the English people have 
diverged in their mental qualities from their Kelto-Teutonic forefathers as far, 
and in the same direction as, their present speech has diverged from its Anglo- 


Algonquian, Iroquoian, Aztec, Mixtec, Quechuan (Peruvian), 
Araucanian, and many of the chief stock languages in every part 
of the New World, while it is not found in any of those of the 
eastern hemisphere. A primordial unity may thus be claimed for 
American speech, which, during the course of its independent 
evolution, shows no clear evidence of having anywhere been 
brought under Asiatic or other foreign influences. It would be 
idle here to discuss the wild statements formerly and even still 
made by erudite etymologists regarding, not merely resemblances 
and affinities, but actual identities between Basque, Irish, Japanese, 
Chinese, Berber, and other tongues of "High Asia/' or of " High 
Africa" on the one hand, and Iroquois, Delaware, Othomi, Maya, 
Peruvian and others of "High America " on the other 1 . All such 
statements are worthless, being based either on the vague and 
unconfirmed reports of "shipwrecked mariners," or on gross 
ignorance of the languages brought into unnatural connection, or 
else on pseudo-scientific processes of comparison incapable of 

Saxon prototype. And then we should have to consider the question of 
miscegenation, to which, as seen on p. 199, race but not language is sus- 

1 One or two instances will suffice to show the reckless nature of some of 
the statements here referred to. In a work on Keltic local names, a fruitful 
source of the wildest etymologies, Herr Obermuller finds Keltic roots referring 
to water in Siberia, India and Peru ; and Prof. John Campbell of Montreal 
has discovered that Creek, Aztec, Choctaw and other American tongues are 
merely so many Japanese dialects. The Abbe Petitot is convinced that 
Athapascan is a disguised Semitic idiom, while Sefior Naxera identifies his 
Othomi (Mexican) mother-tongue with Chinese. Another Mexican, Sefior 
Jose A. Vargas, tells us that the Maxteca language, current on the uplands 
between Puebla and Oaxaca, is identical with that of some gypsies who have 
recently wandered to those parts from the Balkan Peninsula. Hence the 
Maxtecas must be the descendants of other gypsies who came from the same 
region ages ago; for "how can we explain otherwise the fact that the same 
language is spoken in Dalmatia and in these mountains of Mexico" (Monitor 
RepublicanO) Mexico, April 16", 1895). "When I see volumes of this 
character," writes Dr Brinton, "many involving prolonged and arduous re- 
search...! am affected by a sense of deep commiseration for able men who 
expend their efforts in pursuits of such will-o'-the-wisps of science, panting 
along roads which lead nowhere, inattentive to the guideposts which alone 
can direct them to solid ground" (On various supposed relations between 
American and Asian Races > p. 151). 


yielding trustworthy results of any kind. Under critical enquiry 
the linguistic identities between the Old and the New World are 
reduced to the Eskimo dialects current on both sides of Bering 
Strait, where Innuit settlements have long been established. 

Although mainly cast in a common polysynthetic mould, the 
American tongues have, dujpg their separate exist- gur risin 
ence, diverged so widely from the original type that number of 
more irreducible stock languages have been devel- 
oped in this region than in any other part of the 
world. As many as fifty-eight have been determined polysynthetic 
for British North America and the United States e * 
alone, and according to 'some authorities radically distinct lan- 
guages are relatively more numerous in the rest of the continent 
than in the northern regions. Perhaps 150 is not too high an 
estimate for the whole of America, although the researches of 
Buschmann and of some more recent philologists have tended to 
reduce the number of independent linguistic families both in 
Mexico and the United States. Thus the Aztec and the Shoshone 
(Snake) groups would appear to be fundamentally connected, but 
yet so divergent that for the present they must still be treated as 
two independent forms of speech. On the other hand, radically 
distinct languages seem to be less numerous in South America 
than might be inferred from the statements of early writers. On 
the evidence of their speech Mr Clements R. Markham is inclined 
to derive the Amazonian tribes, " now like the sands on the sea- 
shore for number, from two, or at most three parent stocks," 
adding that "the differences in the roots between the numerous 
Amazonian languages are not so great as was generally supposed 1 ." 
Dr Brinton also now abandons the opinion formerly held by him, 
in common with so many other philologists, "that the linguistic 
stocks of South America are more numerous than those of North 
America 2 ." 

Another point of considerable importance is the extremely 
irregular distribution of these stock languages, some of which, such 

1 A List of the Tribes in the Valley of the Amanon, in Jour. Anthrop. 
Institute^ Feb. 1895, p. 336. 

2 The present Status of American Linguistics^ in Memoirs of the Chicago 
Congress of Anthropology, 1893, p. 336. 


as the Athapascan, the Algonquian, the Siouan and the Shoshonean 
in the north, the Nahuatlan and Huaxtecan (Maya-Quiche) in 
the centre, the Guarani and Quechua-Aymara in the south, occupy 
vast areas comparable to those of the Aryan, Ural-Altaic and Bantu 
in the Old World. But the great majority are crowded together, 
like those of the Caucasus and Sudan, in extremely narrow limits, 
as on the north-west coast, where, about thirty are confined to the 
strip of seaboard which extends from British Columbia to Lower 
California between the coast ranges and the Pacific. 
The dassifi- The inevitable result is that classifications have more 

cation of the . . .,,./., 

Aborigines of a linguistic than an ethnical basis ; for how can 

tne most experienced anthropologist pretend to 
distinguish on physical grounds between a few 
thousand Oregon Indians, for instance, who speak a score or so 
of fundamentally distinct idioms, but who all closely resemble 
each other in outward appearance ? As elsewhere remarked (Ch. 
IX.) linguistic are always more easily determined than racial 
divisions, and this is specially the case in the American field, as 
frankly recognised by Mr J. W. Powell, who gives to his valuable 
summary, representing over twenty years' intermittent labours, the 
title of "Indian Linguistic Families of America north of Mexico 1 ." 
For the same reason the accompanying FAMILY TREE OF HOMO 
_ ., - AMERICANUS is necessarily based far more on lin- 

family iree 
of Homo guistic than on ethnical differences. Here Mr 

mencanua. Powell's orthography is adhered to, uniformity in 
this respect being more important than theoretical accuracy. His 
convenient plan of indicating stock languages by the final syllable 

1 Seventh Annual Report of the Bureau of Ethnology ', Washington, 1891. 
Mr E. im Thurn goes even further, and attempts to define ethnical divisions in 
terms of language (op. cit. p. 161). He declares that for Guiana, where 
"there are no very great differences other than those of language," this factor 
"must be adopted" as the basis of classification (#.) and at p. 167 ; "It is 
not very easy to describe the distinguishing physical characteristics of these 
groups [of Guiana natives], for, after all, all being of the same race, the 
differences are but small." Here, it is important to note, the term "race" 
has a very wide meaning, being made commensurate with Homo Americanus. 
It may be added that d'Orbigny's attempt to group the South American 
aborigines according to their physical characters yielded unsatisfactory and 
even contradictory* results (L } Homme Am^ricain^ passim). 



an or tan is also adopted, and extended to the whole of America. 
Thus Siouan, Nahuatlan, Cariban, Guaranian are the collective 
names of families, of which Dakota^ Aztec^ Macusi and Tufi are 
respective sub-groups or branches, confusion being avoided by 
using the plural form where an constitutes an integral part of the 
tribal name, as in Mandans^ Pocomans, Dirians &c. Mr Powell's 
classification is also accepted for North America in all cases except 
the Yuman and Piman groups, which appear as independent 
families on his map, but which are here transferred to the Opatan 
of North Mexico on the authority of Manuel Orozco y Berra, first 
of Mexican systematists ! . 

Assuming a common descent of these multitudinous tribes and 
peoples from more or less generalised Mongol pre- 

Americapro- . , . . . . . . 

babiy peopled cursors in pleistocene and later times, the question 
typiaMUthic, arises b y what route or routes did they reach the 
from Asia by American Continent? It was shown in Chap. X. 

neolithic man. . . * 

that the road by Bering Strait, if not also by the 
Aleutian chain, was always open, and that in late tertiary times an 
alternative highway was probably available from West Europe to 
Greenland and Labrador. It seems likely that both of these 
routes were followed, the western first by primitive long-headed 
tribes, the eastern later by round-headed Mongoloid peoples from 
Asia. That both arrived during the stone ages is evident from 
the presence side by side of the fossil remains of the two types in 
South Brazil and Argentina (p. 98). From the undoubted remains 
of palaeolithic man discovered in the same southern regions 
it would also appear that the long-headed preceded the short- 
headed race, for no clear traces of a round-headed palaeolithic 
people has yet been anywhere brought to light. Thus may be 
explained the presence at the two extremities of the New World 
of highly dolichocephalous peoples, Botocudos, Tehuelche Pata- 

gonians and Fuegian Yahgans in the south. Eskimo 

Present dis- f e , & ' 

tributionofthe in the north from Greenland and Labrador round 
two types. b y the shores of tlie F rozen Ocean to Alaska. These 

first arrivals, being more primitive and armed with ruder weapons, 

1 Geografia de las Lenguas y carta einografica de Mexico^ 1864. 




would, unless absorbed, naturally be driven by the later neolithic 
intruders to the less favoured Arctic and Antarctic regions, where 
their descendants are still found in undisputed possession of their 
uninviting homes. It is noteworthy that "the Eskimo type is 
found in its highest expression in Greenland. Dolichocephaly 
and extreme height of the skull [hypsistenocephaly] become less 
as we approach Bering Strait. The Aleutians and Kolushes 
[Thlinkits] would form the passage between it and the Samoyede 


The Eskimo 

and Mongolian type 1 ." This is precisely what we should expect 
on the assumption of long-headed tribes arriving first from Europe 
and moving westwards till arrested by round-headed 
arrivals from Asia. Doubtless another interpreta- 
tion is given to this fact by Dr Rink and others, 
who trace the Eskimo migrations, not from east to west, but the 
other way, from Alaska to Labrador and Greenland. But all these 
views are based on what may be called local, and consequently 
restricted considerations, which take no account of the broader 

1 Topinard, Anthropology, p. 473. 


and more fundamental factors of the problem 1 . It is not merely 
in the Eskimo domain, but throughout the whole continent that, 
despite the already described secular interminglings, dolichocephaly 
decreases westwards. According to Morton it is in the north 
more prevalent "among the tribes that originally inhabited the 
east of the Alleghanies, and brachycephaly among those to the 
west of the Mississippi. The same thing occurs on the coasts 
of South America 8 ." 

The question is affected or rather obscured by the supposed 
Eskimo affinities observed amongst the Chukchi and other tribes 
in North-east Asia, affinities which are elsewhere explained (Chap. 
XII.). The position here taken is greatly strengthened by the 
comparative study made in Paris of the crania brought by Senor 
Moreno from the Patagonian paraderos, crania which might at 
first sight be taken for "the skulls of Eskimo.... The cephalic 
index is 72*02, that is to say, they are the most decidedly dolicho- 
cephalic in the world after those of the Eskimo [and some 
Melanesians], and their prognathism is 69*4, or less than the 
[normal] American, and as much or more than the Eskimo.... 
This unexpected approximation to the Eskimo suggests some 
curious questions for consideration. Are the Tehuelches the 
autochthonous dolichocephalic element, which by its crossing 
with a race of Asia has given origin to the present American 
type ? May not the craniological singularity of the Eskimo, who 
in certain respects resemble the Samoyedes and the Mongols 
proper, and in others are as distinct as it is possible to be, be 
explained in the same way? They would be another t form of 
cross of the same Asiatic brachycephalic element with the same 
autochthonous American dolichocephalic element 8 ." 

Such an explanation for a polygenist is natural. But by sub- 
stituting quaternary for "autochthonous," which for 
Its solution. o 2 ^ j 

the monogenist has no meaning, M. Topinard's sug- 

1 Chiragh-kt-nlcht ancThtrd, " Under the lamp is darkness," says the 
Hindi proverb. In order to get light on these obscure ethnical problems the 
observer must stand aside, and study them from a distance. 

2 Topinard, ib. p. 480. 

3 Topinard, ib. p. 484; F. P. Moreno, Junr., Des Cimettires et Paraderos 
de Patagonie> in Rev. tfAnthrop. III. 1874. 


gestion will harmonise completely with the solution here proposed. 
The peopling of the New World is thus seen to be intimately 
associated with the seemingly anomalous position of the Eskimo 
and prehistoric Patagonian types, which have hitherto refused to 
adapt themselves to any intelligible scheme, but which now appear 
to fall naturally into their r>lace, to be, in fact, as they are, essen- 
tial elements in the equation. 

One of the most striking, and perhaps the most original, of 
the many alternative theories is that advanced by 
Prof. O. T. Mason 1 , who rightly argues that water theory of the 
yields the easiest means of obtaining food and of America from 
transport, as well as the materials of all the earlier Ind - 

, .' , . Malaysia, 

arts and industries. Hence coastlands, and especi- 
ally estuaries teeming with animal life, first attracted human 
settlers; and on this ground Morgan 2 made the Columbia estuary 
the chief centre of tribal dispersion over the North American 
continent. Following up this line of argument, Prof. Mason 
reasons with much ingenuity that the Columbia river, or some 
neighbouring point, may have been reached at a very remote 
period from Indo-Malaysia by primitive seafarers in rude open 
boats skirting the East Asiatic and North-west American sea- 
bords, and that such voyages may have been constantly made 
thousands of years ago, until the route was interfered with by 
Chinese and other civilised settlers spreading from the interior of 
Asia seawards. Such a route "might have been nearly all the 
way by sea. It could have been a continuously used route for 
centuries. Until interrupted by later civilisations, it might have 
been travelled over for thousands of years. It lies absolutely 
along a great circle of the earth, the shortest and easiest highway 
upon a globe" (p. 279). Reference is made to the analogous 
case of the British Columbian Haida Indians, who for ages have 
annually voyaged in their frail craft five hundred miles southwards 
to Puget Sound in quest of clams and oysters for their own con- 
sumption and for trade. Weight is also placed on assumed 

1 Migration and the Food Quest, A Study in the Peopling of America, 
reprinted from The American Anthropologist for July 1894; Washington, 

2 North American Revieiv, Oct. 1869, Jan. 1870. 


ethnical and possible linguistic affinities along the line of primeval 
traffic ; on the favourable marine and aerial currents ; on similar 
social institutions, arts and industries of too striking a nature to 
be explained otherwise than by actual contact. It is asserted that 
scarcely an original idea, not even the game of patolli 1 , "was 
developed upon the western hemisphere" (p. 290); that "this 
close connection between the two continents has existed for 
thousands of years," and that " there never was known to history 
a day when the two continents were not intimately associated " 
(p. 292). 

The case could hardly be put in stronger language, and, if it 

NC ative could be upheld, many pages of this work would 

objections to have to be re-written. But it may be asked, if history 

this theory. ^ ^^ a j wa y s ^ een j n touc h w ith t h e New World, 

why did the New World need to be discovered by Columbus, or 
his Norse precursors ? And if this close connection existed " for 
thousands of years," how did it happen that there was no inter- 
change of the useful commodities of social life between the two 
hemispheres? These should have preceded, or at least accom- 
panied, the aesthetic fancies assumed to have been wafted over the 
seas from Malaysia or Papuasia to the north-west coast of America. 
But while this region received none of the good things of the 
East, neither its silks, iron, cereals such as wheat, rice, and millet; 
pulse such as pease and lentils; nor its beasts of burden such as 
the horse, ass, and camel, on the other hand none of the fruits of 
the West, maize, tobacco, potatoes, tomatoes, and the like, found 
their way to the East, so that after thousands of years this inter- 
national traffic produced nothing but negative results, hence 
might as well never have existed. But interchange is the very 
essence of commercial intercourse ; therefore the assumption falls 
to the ground, the more so that history knows nothing of this 

1 As much has been made of the undoubted resemblance between this 
Mexican game, and the pachesi, a kind of backgammon long known in India, 
it should be stated that, after a careful study of the subject, Mr Culin and 
Mr Frank Gushing declare patolli to be "thoroughly American in origin." 
(See Dr Brinton, On various supposed Relations between the American and 
Asian Races > p. 149.) The question was first raised by Dr E. B. Tylor 
(Jour. Anthrop. Inst. VIII., 1878). 


close connection between the two hemispheres. The only inter- 
course known to (recent) history is that which has long been 
carried on between the Eskimo and other tribes on both sides of 
Bering Strait, an intercourse which has led to some ethnological 
mystifications, but which leaves the question of early intercon- 
tinental migrations untouchjd. 

Nor could such migrations be explained by Prof. Mason's 
theory, even were the main facts admitted. It 
would not account for the presence of two types of 
primitive man in the southern extremity of the New 
World in quaternary times, that is, ages before the development of 
navigation or of any other advanced art in Malaysia or Indonesia. 
Primitive man did not reach America from those regions by 
water, but from Asia and Europe by the overland routes, as 
explained. But it is not a question of primitive man, but of" East 
Indians," and "Malays" (p. 281), and of " pre-Malays, who were 
the Phoenicians of the Orient " (p. 255), that is to say, cultured 
peoples, who had long outlived the stone ages. If therefore these 
were the first settlers in the New World, what becomes of the 
American palaeolithic man ? And if he be " discredited," there 
is still the American neolithic man, accepted by all, but unac- 
counted for by this theory. Did these " Phoenicians of the Orient " 
revert in America to the savage state, settle down on the shores 
of New England, Brazil, and Fuegia and build up the enormous 
kitchen-middens of those regions ? Did they fabricate the multi- 
tudes of rude stone implements which have been collected in tens 
of thousands from all parts of the United States (p. 105), and which 
cannot be distinguished in form from the European palaeoliths? 
Did they build the mounds of the Ohio valley, the casas grandes 
of the Pueblos, the Mexican teocalli, the great cities of Yucatan 
and Peru, the megalithic monuments of Lake Titicaca? Did they 
forget their Malayo-Polynesian and other eastern tongues, and 
invent new forms of speech in the New World, forms utterly 
unlike any current in the Old ? Surely all these things should be 
taken into account in any rational theory that may be advanced 
to explain the origin of the American aborigines, and their orderly 
evolution up to the various planes of culture reached by them in 
pre-Columbian times. 


Thanks to their generally homogeneous character, and to their 
independent normal development since the stone 
a g es * n an environment separated from the rest of 

tain usages the globe, the American aborigines present few 

and mental , , ^^ ^ 

aspects of the other racial problems of sufficient importance to re- 
discussion in thesfe pages. Once severed from 

New Worlds fa e fictitious Asiatic connection and influences, 


the study of their social, religious, and political 
institutions acquires quite an exceptional interest. Striking re- 
semblances and points of apparent contact with the usages of 
the eastern populations at corresponding grades of culture are no 
longer to be explained by the clumsy device of importations, 
impossible borrowings or affinities, but by the immeasurably 
more rational conception of their common mental constitution. 
Such coincidences thus become doubly instructive. They not 
only illustrate the social condition of the peoples themselves, 
but also throw a flood of light on the primeval psychic character 
of all mankind, as clearly appears from the all-embracing but 
unfortunately somewhat entangled ethnico-psychological writings 
of Dr Adolf Bastian. Thus, to give one instance amongst a thou- 
sand, instead of deriving Papuans from Basques, and Basques from 
Guiana Indians, because of the couvade common to all, it will be 
more profitable to study the motives and mental processes which 
underlie that strange custom, and which may explain its inde- 
pendent origin amongst such widely separated and fundamentally 
distinct peoples. By adopting this course, Mr James Rodway 
seems to have arrived at a rational solution of the mystery. On 
the birth of the child, the father "calmly prepares to do what he 
considers his duty. He must not hunt, shoot, or fell trees for some 
time, because there is an invisible connection between himself 
and the babe, whose spirit accompanies him in all his wanderings, 
and might be shot, chopped, or otherwise injured unwittingly. 
He therefore retires to his hammock, sometimes holding the little 
one, and receives the congratulations of his friends, as well as the 
advice of the elder members of the community. If he has occasion 
to travel, he must not go very far, as the child spirit might get 
tired, and in passing a creek must first lay across it a little bridge, 
or bend a leaf in the shape of a canoe for his companion. His 


wife looks after the cassava bread and pepper-pot, and assists the 
others in reminding her husband of his duties. No matter that 
they have to go without meat for a few days, the child's spirit 
must be preserved from harm 1 ." So with the Egyptian and 
American pyramids, on which so many wild theories have been 
based, but all of which are^independent local developments ori- 
ginating in the same psychic feeling, awe and fear of the dead. 
They must be honoured with parting gifts; their remains and 
belongings, deposited in cists, must be guarded against profanation 
by superimposed mounds (p. 128); their wrath must be appeased 
by periodical offerings and by sacrifices on their graves. Hence 
the mounds may in some places assume a truncated form for 
the convenient celebration of these rites, and for the erection 
of permanent buildings for the same purpose. Thus arose the 
" temple-mounds" of the Mississippi basin described by Mr Lucien 
Carr 2 , and the Mexican and Maya teocalli, all of which, like the 
Egyptian pyramids, contain human remains, but none of which 
can date farther back than about the sixth century of the new era, 
that is to say, ages after pyramids had ceased to be built by the 
Egyptians, to whom, nevertheless, these American structures have 
been attributed by those who refuse to credit the natives of the 
New World with a single original idea. 

It would be surely more reasonable to attribute the "temple- 
mounds" to the vanished race, by whom somewhat analogous 
monuments were raised in Tahiti, the Low Archipelago and 
other South Sea islands. "In the Society Islands, as in many 
other parts of the Pacific, are to be found a number of buildings 
which testify to the existence in former times of a people 
of a higher development. They are generally in the form of 
terraces or platforms, placed in elevated spots, and formed of 
hewn blocks of stone which are often of great size. In the 
centre is placed a sort of massive altar. A very large building 
of this kind exists at Papawa in Tahiti. From a base measuring 
270 feet by 94 feet rise ten steps or terraces, each about 
6 feet in height. The object of these morals, as they are termed, 

1 In the Guiana Forest ', 1895, pp. 25, 26. 

3 The Mounds of the Mississippi Valley historically considered; Smithsonian 
Report, 1891, pp. 95 et seq. 

K. 24 


is not very clear. They were in many cases no doubt of a 
monumental, if not sepulchral, nature; but sacrifices were ap- 
parently offered upon them in some instances, and it seems that 
they also served on occasions as forts or strongholds 1 ." Here 
the tables might well be turned on those archaeologists who 
trace the foundations of every monumental structure in the New 
World to the Eastern Hemisphere ; for it might be argued that, if 
the Egyptians built, for instance, the pyramid of Cholula, which, 
like that of Cheops, "may have been a tomb" (de Nadaillac, 
p. 351), the morai of Papawa may & fortiori have been erected 
by the Toltecs, or any other prehistoric cultured people of 
Central America, the resemblances between the morais and the 
terraced Mexican pyramids being so much greater than that 
between these structures and the pointed pyramids of the Nile 
Valley. But all such inferences are highly unscientific, and it 
may be confidently asserted that, if Cholula were of Egyptian 
workmanship, the proofs would lie on the surface as palpably as 
the proofs of Hindu influences lie on the surface of Boro-bodor 
and Angkor- Vat. It may be concluded with Mr Cyrus Thomas 
that, "the mind and requirements of man being substantially the 
same everywhere and in all ages, the primitive works of art which 
relate to supplying these requirements will be substantially the 
same where the conditions are alike" {Mound Explorations, 1894, 

P- 5 2 9)- 

The fate of the aborigines since the discovery of America has 
been compared by Dr Daniel Wilson with that of 

Results of . . . 

the discovery the men of the Stone Ages in Europe, when their 
ment^of *Am*- domain was invaded by "one or more races superior 
rica on the alike in physical type and in the arts upon which 
progress depends 2 ." But, owing to the different de- 
grees of culture prevailing in America, the results have not every- 
where been the same. The normal development of the leading 
nations Aztecs, Mayas, Chibchas, Peruvians who had established 

powerful political systems with thoroughly organised governments, 

1 Dr F. H. H. Guillemard, Australasia, p. 515. 

2 American If lust rations of the Evolution of New Varieties of Man> Jour. 
Anthiop. Inst. t 1878, p. 340. 


was abruptly arrested, and replaced by the social and religious 
institutions of the conquerors. Millions perished during the first 
conflicts, and later in the mines or on the planta- in Latin 
tions. In the West Indies all the natives rapidly America - 
disappeared, and here their place was taken by negro slaves 
imported from Africa. But^ elsewhere the civilised populations 
survived in sufficient numbers to amalgamate with the Spanish 
and Portuguese intruders, and form the substratum of the present 
mixed peoples of Latin America (p. 152). 

In the northern continent totally different conditions produced 
totally different results. Here the normal relations Jn An lo _ 
of a few hundred thousand half-savage and partly Saxon 
agricultural hunting tribes, distributed over several 
million square miles of territory, were at first little affected by a 
few British settlements on the eastern seabord, mostly engaged 
in hostilities with rival French colonists in the St Lawrence basin. 
Spanish America was overrun and largely reduced within a single 
generation after the fall of Mexico, whereas the Prairie Indian 
was still roaming the Mississippi plains far into the nineteenth 
century. On the other hand no fusion of the two elements has 
taken place in Anglo-Saxon America at all comparable to the 
amalgamation of Europeans and natives in the central and southern 
regions. Here the union has been reciprocal, equally affecting 
both races, whereas in the north it has been, so to say, one-sided. 
It was shown (p. 152) that the North American Indians have almost 
everywhere received a strain of white blood ; but the white popu- 
lations, always excepting the French Canadians, have on the 
whole preserved their racial purity intact. In virtue of a deeply- 
rooted ethnical sentiment, the half-breeds have, as a rule, failed to 
acquire citizenship amongst the higher race, and are fain to cast 
in their lot with the aborigines who are now for the most part 
confined to reservations. Recently, however, a tendency towards 
absorption in the white population has been observed in some of 
the western states, but always under the indispensable condition 
of tribal effacement. "There is one way," writes Mr James 
O. Dorsey, "in which a diminution of some tribes is taking 
place, viz. by ceasing to be Indians and becoming members of 
civilized society. In Minnesota all persons of mixed blood, 

24 2 


i.e. of white and Indian descent, are recognized as citizens. The 
same is true in other States; and the privilege is extended to 
those who are not mixed bloods. Also, under present homestead 
laws, Indians are becoming citizens by going off their reserves. 
Let a well-arranged severally bill be enacted into a law, and 
Indians be guaranteed civil rights as other men, and they will 
soon cease to be Indians 1 ." 

But absorption, universal in Latin America, is still the excep- 
tion in the United States, where the natives are consequently 
doomed to almost absolute extinction. At least the slight Indian 
strain that may survive amid the white populations may be re- 
garded from the ethnical standpoint as une quantite n'egligeable. 
Even Dr Wilson, who is perhaps inclined to exaggerate the import- 
ance of the aboriginal element, admits that "the red race is actually 
disappearing by positive extinction," adding, however, that "it is 
blending by a process of absorption into the dominant race, not 
without leaving some enduring influence on the European-American 
population both of Canada and the United States " (loc. cit. p. 356). 
Although in the States this " influence " must be regarded as 
infinitesimal, some ethnologists have nevertheless 
American g " attributed to it a certain approximation to the Indian 
to P m?s"e' en* physical type, which has been observed amongst 
ation, but to the white populations, especially in some of the 
convergence. sou thern and central states. But this approxima- 
tion, which reveals itself in the increased stature, slender and 
somewhat bony figure, sharp angular features, pale or less florid 
complexion, straight and stiff black hair 2 , is certainly not due to 
crossings with the aborigines, for similar tendencies have already 
been developed amongst the British settlers in Australia. It is 
to be regarded rather as a case of convergence, such as that of 
the Germans in Trans-Caucasia (p. 203), and may be attributed 
to the changed climatic conditions, drier, hotter and less nebulous 
than those of the British Isles. But there can be no question of 

1 Contributions to North ^American Ethnology, IX. Washington, 1893, 
p. 167. 

2 The long lank hair " is, in comparison with the soft silky hair of the 
Englishman, evidently an approach to the American Indian" (Waitz, An- 
thropology^ p. 54). 


degeneracy of the Anglo-American populations in their new 
environment. The lugubrious vaticinations of a now-forgotten 
school of fierce polygenists have already been belied by the 
magnificent physique of the Kentucky and Tennessee peoples, 
mainly sprung from a hale Virginian stock, with no appreciable 
strain of fresh blood from the mother country. 



North Africa probable cradle of the Caucasic race which spread thence east 
to Asia and north to Europe The Cro-Magnon and other early European 
races affiliated to the fair Berbers of Mauritania West Europe occupied 
by several varieties of Homo Caucasicus in the Stone Ages Who were of 
non- Aryan speech like the still surviving Basques The I bero- Berber 
problem Basques and Picts Family Tree of Homo Caucasicus 
Xanthochroi and Melanochroi Blacks of Caucasic Type Physical 
Characters of Homo Caucasicus White, Brown and Dark Hafnites 
The Tamahu Karaites of the Egyptian records The "New Race" in 
the Nile Valley The Eastern Hamites : Afars ; Bejas ; Gallas and 
Somals; Masai and Wa-Huma Ethnical relations in Abyssinia: Him- 
yarites ; Agaos ; The present Abyssinian populations Relations of the 
Hamites to the Semites The Semitic Domain The Semitic Groups-- 
Semitic physical and mental characters The Semitic Languages The 
Aryan-speaking Peoples Aryan a linguistic not a racial "expression 
True character of the Aryan migrations Illustrated by the Teutonic in- 
vasion of Britain; and by the Hindu invasion of India The Aryan 
Cradleland Primitive Aryan Culture Schrader's hypothesis Conflicting 
views regarding the Aryan Cradleland reconciled The Eurasian Steppe 
true home of the primitive Aryan Groups The primitive Aryan type 
difficult to determine But probably xanthochroid The Aryan problem 
summed up Recent expansion of the Aryan-speaking Peoples The 
"Greater Britain'* The Aryan linguistic family Table of the Aryan 
linguistic groups Disintegration of primitive Aryan speech The Teu- 
tonic phonetic System Ethnical and linguistic relations in the Caucasus 
Main Divisions of the peoples and languages of Caucasia Ethnical 
and linguistic relations of the Dravidas Sporadic Caucasic Groups: 
Todas ; Ainus. 

FOR the history of primitive man in the northern hemisphere 
the chief geological factor is the condition of the 

North Africa ,.. ... T - -, , , 

probable cradle Mediterranean basin in miocene and later epochs. 

Reiference has already been made to the distribution 
of land and water after the slow disappearance of 

the miocene continent, and it will suffice here to add that Prof. 

E. Hull has lately placed beyond reasonable doubt the existence 


of barriers, by which the Mediterranean area was separated into a 
chain of basins in post-miocene times 1 . Continuous land, or at 
least land connecting North Africa, Europe and West Asia at 
several points during the pliocene and post-pliocene epochs, is 
thus established, and at once explains the constant migrations 
of the large African fauna north and south of the Mediterranean 
basin. That these migrations were accompanied by primitive 
human groups is sufficiently attested by the overwhelming proofs 
of their presence on both sides of this area during the Stone Ages. 
The long sojourn of palaeolithic man in Mauritania, using the term 
in its wider sense, has been revealed by the researches of Dr 
Collignon and of Dr Couillault in the Gafsa district, Tunisia 
(p. 92), and it was also seen (pp. 134-5) that the same region was 
one of the earliest, and in every respect one of the most important 
centres of neolithic culture. Human progress, arrested or at least 
partly interrupted in the north by the phenomena of glaciation, 
subsidence, and upheaval, was exposed to none of these disturbing 
influences in the south, where the Sahara itself formed a well- 
watered and habitable region, and not, as commonly supposed, a 
marine bed. Here therefore pliocene man, migrating from his 
original seat in the Indo- African Continent (Chap. X.), found a 
new home where by slow adaptation to the changed and im- 
proved climatic conditions the highest human type, conventionally 
known as the Caucasic, may well have been evolved. The white 
man and the negro, says a great biologist, have been differentiated 
"through the long-continued action of selection and environ- 
ment 2 ." 

From this centre of evolution and dispersion the higher groups 
passed by easy transitions, eastwards into the Nile 
valley 8 and West Asia, northwards to Iberia, and 

thence to West and Central Europe. But these Asia and north 

v . to Europe. 

migrations, like those of the African fauna itself, 

1 Paper read before the Geological Society, Feb. 6, 1895. 

2 The late Prof. Arthur Milnes Marshall, Biological Lectures, 1894, pp. 247 
and 350. 

3 Thus M. G. Maspero holds that the Egyptian people presents the 
characteristics of those white races which have been found established from all 
antiquity on the Mediterranean slope of the Libyan Continent. " This popula- 


were successive and spread over a vast period of time, during 
which the process of upward physical and mental development 
was in continuous progress. Thus is explained the appearance of 
low human types (Neanderthal, Spy, Castenedolo) in various parts 
of Europe during late pliocene and early pleistocene times. They 
represent the first waves of migration from North Africa soon 
after the arrival of pliocene man in that region. But they were 
followed later by higher types, such as that of Cro-Magnon, 
which radiating from the Vezere district, gradually 
Magnonand spread over a great part of Europe, and is by 
EurJp C eYi y some ethnologists already regarded as the substra- 
raccs affiliated turn o f the present populations of West Europe. De 

to the fair Ber- ^ r , f . u f r 

bers of Mauri- Quatrefages does not hesitate to connect all the fossil 
tama * remains found in Europe with "the white type 1 /' 

and if these remains be regarded as so many transitional forms in 
the evolution of Homo Caucasicus, there can be no objection to 
that view. He also agrees with M. Verneau in identifying the 
Cro-Magnon race with those groups of tall, dolichocephalic 
Kabyles (Berbers) of fair complexion and often characterised by 
blue eyes, who still survive in various parts of Mauritania, and 
were even represented amongst the Guanches of the Canary 
Islands (p. 446). 

But Jn consequence of his hypothesis of a northern origin of 
Homo Sapiens, De Quatrefoges is obliged to introduce the Cro- 
Magnon race apparently from Siberia, " arriving in Europe simul- 
taneously with the great mammals which were driven by the cold 
from Siberia, and no doubt following their route" (#.). Thus 
their later migrations are described as following a southerly course, 
from Belgium, France, Iberia and Italy to North Africa and the 
Canaries. But the movements of the great mammals were not 
from north to south, but to and fro, over the Eurafrican Continent, 
for this fauna was essentially southern, and advanced and retreated 
synchronously with the advance and retreat of the ice sheet. 
Hence it is that this exceedingly diversified fauna is scarcely 

tion is of African origin, and came to Egypt from the west or south-west " 
(The Dawn of Civilisation Egypt and Chaldaa, English ed. by M. L. McClure, 

1 Op. cit. p. 441. 


represented in Siberia except by the rhinoceros and mammoth, 
whereas the early and later species of elephant, lion, bear, 
hyaena, and hippopotamus abound in Britain, France, Italy and 

Thus all the conditions point to North Africa as the true 
centre of dispersion for the jliocene and pleistocene 
mammals, which invaded Europe in successive occupied by PC 
waves of migration during those epochs, and which several varie- 

. , . ... . ties of Homo 

were admittedly accompanied by primitive man in Caucasicusin 
ever increasing numbers. So thickly inhabited had 
some more favoured districts become in later, but 
still remote times, that the human remains brought tb light by 
M. de Baye in the neolithic caves of the Marne basin already show 
an intermingling of no less than " six races, representing at least 
three quite distinct types, with an aggregate of characters and a 
physiognomy which closely recall what may be seen in the most 
modern craniological collections" (ib. p. 441). As these remains 
are all connected with the "white type" (see above), it follows 
that several varieties of Homo Caucasicus were already developed 
in neolithic times in West Europe 1 . It was suggested (p. 136) that 
none of these pre-historic peoples were of Aryan speech, from 
which a fresh argument may be drawn in favour of their arrival 
from North Africa, where no Aryan language was ever current 
before the Greek occupation of Cyrenaica (yth century B.C.). In 
this connection the importance of the survival of a non-Aryan 
form of speech, still spoken by the Basques on both sides of the 
Western Pyrenees, can scarcely be overrated. The significance of 
this fact is greatly increased since modern research tends more 
and more to connect both the Basque people and their primitive 
language with the indigenous Hamitic (Berber) race and language 
of Mauritania. We have seen (p. 205) that the late G. von der 
Gabelentz claims to have established a connection between the 
Basque and Berber linguistic groups. A similar connection 

1 The same inference is drawn by Prof. Kollmann from a study of the 
neolithic remains in the Swiss barrows, "welche zeigen aufs Neue dass die 
Lang-, wie die Breitgesichter von uralter Herkunft sind und schon damals 
verschiedene Varietaten neben einander lebten" (Zeitsch* /. Ethnologic* 1894, 
Heft v. p. 221). 


between the Basque and Berber physical types has long been 
The ibero proclaimed by French and Spanish anthropologists, 
Berber pro- and although a distinct Basque type has lately been 
denied 1 , it has nevertheless been, so to say, recon- 
stituted by the recent measurements of Basque conscripts taken 
on the French slope of the Pyrenees * 9 These measurements fully 
confirm the views of Dr. F. M. Tubino 8 regarding the identity of 
the Basques with the ancient Iberians, and their relationship to the 
fair Berbers of Mauritania, as well as to the fair Libans (Libyans) 
depicted on the Egyptian monuments of the i4th and I5th cen- 
turies B.C. It is also to be noticed that the megalithic monuments 
of Iberia, which abound especially in western Andalusia, in 
Portugal, Galicia and generally along the north coast, recall 
"rather the megalithic monuments of Northern Africa than those 
of Brittany and of the British Isles 4 ." But despite local differ- 
ences, which characterise all wide-spread cultures, it has already 
been pointed out (Ch. VI.) that all these neolithic monuments 
were erected by the same race, by whatever name they be called 
Berbers and Libyans in Africa, Iberians and Turdetani in Spain, 
"Kelts," "Gauls/ 7 Picts in Gaul and Britain. This view is con- 
firmed by the researches of Prof. John Rhys, Mr J. Gray and 
others, who are now disposed to give a wide expansion northwards 
to the Iberian race, identifying them with the Picts, that is, the 
Pictones of Poitou, and the indigenous Pictish inhabitants of the 
British Isles. Prof. Rhys certainly draws a distinction between 
Picts and Basques; but he supposes them to be "as 
nearly related to one another as Latins, Teutons 
and Kelts are held to be related within their own 

1 "II n'y a point de type basque" (Elisee Reclus, I. p. 855). 

2 Thousands of French Basque recruits have been examined by M. R. 
Collignon, who establishes a Basque type specially characterised by "le 
renflement du crane au niveau des tempes, et le prodigieux retrecissement 
de la face vers le menton," while in several respects recalling the features 
of the ancient Egyptians and Berbers (La Race Basque in V Anthropologie^ 
July 1894, passim). This anthropologist admits a difference between the short- 
headed French and the long-headed Spanish Basques, but holds that the 
French represent the purer type in every respect. 

8 Los Aborigines Ibericos. 

4 Wentworth Webster, Academy, Sept. 26, 1891. 


Aryan family....! believe Picts and Iberians to have belonged to 
one and the same family, which I have ventured to call Ibero- 
Pictish. How nearly related Picts and Iberians may prove to be 
is a matter for future research 1 ." But Mr Gray seems needlessly 
to separate the Basques from the Iberians, and to connect the 
former positively with the IJicti of North Britain and the Pictones 
or Pictavi of South [West] Gaul. " The language of the Picts was 
Basque. The name Pict is derived from a Basque word, pikatu, 
to cut.... The pre-Pictish inhabitants were probably Iberians, and 
prevailed mostly in Ireland, South Wales, Cumberland and South 
Scotland 2 ." It is right to add that these conclusions are far from 
being accepted by some of the leading Keltic scholars, such as 
Mr Whitley Stokes and Prof. Windisch, both of whom still hold 
that the Picts were Kelts, " but more nearly allied to the Cymry 
[Welsh] than to the Gael [Irish] 3 ." 

But these discordant views on points of detail do not affect 
the main argument, that Homo Caucasicus had his ., - 

.... . Family Tree 

"origin in North Africa, and spread thence in palaeo- of Homo Cau- 
and neolithic times over the whole of Europe, the 
Nile Valley and a great part of Asia. In the accompanying Family 
Tree are seen the chief branches, which have ramified from the 
parent stem during pre-historic and historic times. 

In all attempts at a classification of Homo Caucasicus, claim- 
ing to be something more than a mere linguistic xanthochroi 
grouping, the great initial difficulty is colour. So and Meiano- 
true is this that, as seen, Huxley and other recent 
systematists begin at once by splitting the whole division into two 
sections, a fair and a dark type the Xanthochroi and the Melano- 
chroi branches of our Family Tree. But even this is far from 
covering the whole ground. It not only leaves out of account the 
widespread Indonesian branch, here ramifying to the left, which is 
neither fair nor dark, but distinctly brown, but it also gives to the 
term "dark" a totally inadequate meaning. Melanochroi in fact 

1 Academy ', Sept. -26, 1891. 

2 Distribution of the Picts in Britain, as indicated by Place-Names^ Paper 
read at the Meeting of the Brit. Assoc. Oxford, 1894. 

3 W. Stokes, The Linguistic Value of the Irish Annals. 



is not taken in its strict sense at all 1 , having reference not to a 
" black," but to a " pale " colour of the skin usually accompanied 
by black hair and eyes : " West of the area occupied by the chief 
mass of the Xanthochroi, and north of the Sahara, is a broad belt 
of land, shaped like a Y- Between the forks of the Y nes tne 
Mediterranean, the stem <jf it is Arabia... The people inhabiting 
the area thus roughly sketched have, like the Xanthochroi, pro- 
minent noses, pale skins, and wavy hair, with abundant beards; 


(Xauthrochoid Type.) 

but, unlike them, the hair is black or dark, and the eyes unusually 
so. They may thence be called Melanochroi '. . .They are known 
as Kelts, Iberians, Etruscans, Romans, Pelasgians, Berbers, Sem- 
ites. The majority of them are long-headed, and of smaller 
stature than the Xanthochroi 2 ." But within the Caucasic division 
there are several groups, such as the eastern Hamites (Bejas, 
Agaos, Somals, Gallas), and the Abyssinian Semites 
(Tigre, Amhara), besides many Hindus and Dravi- 
das, who have not merely black hair and eyes, but 

1 "Black-hued," from Gr. juAas, black and xP ot( *> colour. 

2 Huxley, Critiques, p. 151. 


also very black skins. Some even of the Sudanese Arabs, 
notably the Sheygyeh people between Dongola and Abu-Hammed, 
are remarkable for their extremely dark complexion, although 
claiming pure Arab descent. So also the Dazas, or Southern 
Tibus of the Sahara north of Lake Chad, and the Harratin, or 
" Black Berbers " of Tidikelt and tj?e Saharan oases, many of 
whom are blacker than the average Negro. 

But it may be asked, on what ground are these dark groups 
_ . , . included in the light-coloured Caucasic division, 

Physical cha- & ' 

ractersofHomo where their very presence seems to involve a con- 
tradiction in terms? The reason is, because they 
cannot be separated anthropologically from that connection. 
Apart from the colour, which in some cases appears to be the 
result of climate and in others is certainly due to an infusion of 
Negro blood, these " black Caucasians," if the expression can be 
tolerated, are amongst the very finest representatives of the 
Caucasic type. According to Messrs Flower and Lydekker, this 
type is distinguished generally by light skin, though in aberrant 
groups as dark as the Ethiopic ; hair ranging from fair to black, 
soft, straight or wavy, in transverse section intermediate between 
the flat Ethiopic and round Mongol ; full beard ; skull variable, 
though mostly mesocephalic ; jugal bones retreating ; face narrow 
and projecting in the middle line (pro-opic) ; orbits moderate ; 
nose narrow and prominent (leptorrhine) ; jaws orthognathous ; 
teeth small (microdont) l . With regard to Huxley's blonde and 
dark divisions, these anthropologists hold that, despite differences 
of colour of eye and hair, they agree so closely in other respects 
that they are best regarded as modifications of one great type than 
as primary divisions. In any case they are now mostly blended 
together in diverse proportions, and even the blonde, though 
found chiefly in North Europe, extends to North Africa [where in 
fact it originated] and eastwards to Afghanistan. In this careful 
survey of the whole field, the dark division receives its full expan- 
sion, comprising not only black hair and eyes, but also a skin of 
almost every shade from white to black (p. 753). 

There is thus no reason to create separate divisions for all 

1 Op. tit. p. 746. 


these groups, which possess so many physical cha- White brow 
racters in common. To do so would lead to nothing and dark 
but confusion, as, for instance, in the case of the Hamites> 
various Berber groups, all agreeing in their fundamental features, 
although some may be black, some brown or swarthy, some fairer 
than many Europeans. The black Harratins of the southern 
oases have for neighbours and kinsfolk the Kabyles of the Mauri- 
tanian uplands, "many of whom have a fair complexion and 


(West Hamitic Type.} 

blonde hair, recalling the peasantry of North Europe rather than 
the inhabitants of Africa 1 ." Even the Arabised Berbers of North 
Morocco are described by Mr Walter B. Harris as "for the most 
part fair, with blue eyes and yellow beards, perfectly built and 
exceedingly handsome men 2 ." Such features have been attributed 
to contact with the Roman colonists, and even to the Vandals, 
who invaded and occupied the whole of Mauritania in the 5th 
century. But the Periplus bearing the name of Scylax (Herodotus 
iv. 44) already mentions a people of fair complexion on the shores 

1 M. Shaler, Esquisse <le FEtat d'Algcr, p. 119. 

2 Proc. R. Geograph. Soc. 1889, p. 490. 



of the Lesser Syrtis, and the Tamahu 1 Libyans are figured on the 
Egyptian temples (1500 1300 B.C.) with a rosy skin, blue eyes 
and red or light hair. Similar traits occur even amongst the 
The Tamahu Tuareg Hamites of the Sahara, who are not known 
Hamitesand to have ever had direct relations either with the 
Race "in Romans or the Vandals. In a word, these Berber 

Egypt. populations, forming the true indigenous element 

throughout North Africa, are essentially Europeans fl . They were 


(Berber Type.) 

not merely the allies, but the kinsmen of those blue-eyed and 
light-haired peoples (Pelasgians, Teucrians, Hellenes, Itali, Etrus- 

1 This word still exists under various dialectic forms (TamaMg, Tamashek, 
Tamazigt] applied collectively to the Hamitic languages of the Sahara and 
Mauritania. The form T-amazig-t, when stripped of its fern, prefix and 
postfix particle /, is seen to be identical with the Maxyes of Herodotus (later 
Masices, Mazices), i.e. Amzigh, pi. Imazighcn, " Freemen," the most general 
name of the Mauritanian Berbers. 

2 " Les Berbers de 1' Atlas, en eflfet, et rneme la generalite des Touareg..., 
sont physiquement de veritables Europeens... Compare a 1'Arabe, ou I'Euro- 
pe>n, le Berber a des differences de physionomie, non des differences de type " 
(V. de Saint-Martin, Nouveau Dictionnaire etc., I. p. 411). 


cans ?), who in the time of Ramses II. descended from the islands 
of the Mediterranean on Lower Egypt, and who were expelled by 
Ramses' son and successor, Sethi II. of the ipth dynasty. 

But long before the iQth dynasty a large part of Egypt had 
been occupied by a people of fair type, who held possession of a 
tract over a hundred miles in length between Abydos and 
Gebelen during the yth, 8fli and 9th dynasties, that is to say, 
about 5000 years ago. This is the so-called "new race," whose 
arts, industries, graves and osseous remains were unexpectedly 
brought to light in large numbers by Mr Quinbell and Prof. 
Flinders Petrie in 1894-5. "The race was very tall and powerful, 
with strong features : a hooked nose, long-poirited beard and 
broivn wavy hair are shown by their carvings and bodily remains. 
There was no trace of the negro type apparent, and in general 
they seem closely akin to the allied races of the Libyans and 
Amorites. ...Though some objects point strongly to an Amorite 
connection, others indicate a western source ; and it must be 
remembered that probably the Amorites were a branch of the 
fair Libyan race. The geographical position is all in favour of 
the race having come into Egypt through the western and great 
Oases; for the yth and 8th Egyptian dynasties were still living at 
Memphis, showing that no people had thrust themselves up the 
Nile valley 1 ." On one of the skulls in the collection of objects 
exhibited at University College, London, in 1895, the hair still 
adheres to the scalp; it is of a darkish, almost russet-brown hue, 
and very curly like that common amongst the Hellenes and other 
South Europeans. The "new race" must clearly have been a 
people of fair Caucasic type, probably of the same stock as the 
ancient Ibero-Libyans, that is, the above-mentioned Tamahu of 
later Egyptian documents. If so they are still perhaps best 
represented by the fair blue-eyed Berbers of Mauritania, and, 
despite their antiquity of some 5000 years, mark a relatively late 
stage in the eastward spread of the primitive Caucasic peoples 
from their North African cradleland. They thus afford un- 
expected confirmation to the views here advocated regarding 
Caucasic origins and early migrations through Egypt eastwards to 
Asia and southwards to Ethiopia. Gebelen, southernmost knowa 

1 H. M. Flinders Petrie, Academy^ April 20, 1895, p. 342. 
K. 2 


limit of their territory, lies not far from the frontier of Nubia, a 
region which since Roman times (Diocletian) has been occupied 
by Negroid tribes from Kordofan, but which had at an earlier 
period been held by the cultured Hamitic Blemmyes. It is note- 
worthy that these Blemmyes of /Ethiopia supra &gyptum, re- 
garding whose affinities much doubt had long prevailed, are now 
regarded by Prof. Sayce as " of Berber race and language. Prof. 
Maspero has shown (in the Transactions of the Society of Biblical 
Archeology r , Vol. I., p. 127) that in the time of the nth dynasty 
a particular species of dog was called in Egyptian by the foreign 
name of abakru, which is the Berber aba'ikur *a dog,' from which 
we may infer that a Berber language was already spoken in the 
neighbourhood of Thebes. Herodotus (n. 42) asserts that the 
inhabitants of the Oasis of Ammon, the modern Siwah, were a 
mixed colony of Egyptians and Ethiopians ; and since a dialect is 
now spoken there akin to those of the Tuaregs and Kabyles, it 
would seem that these Ethiopians were a Berber tribe.... If my 
arguments are sound, we shall thus have to look to the Berber 
languages for an explanation of the Meroitic inscriptions 1 ." The 
almost simultaneous researches of Prof. Sayce and of Prof. Petrie 
in the Nile Valley thus complement each other. They attest in 
the whole of that region the presence at a remote epoch of 
Hamitic Ethiopians and Libyans, and explain the juxtaposition 
of these two peoples in the Second Book of Chronicles, where it 
is asked, " Were not the Ethiopians and the Lubim a huge host, 
with very many chariots and horsemen ? " (xvi. 8). 

But, as already pointed out, M. Maspero holds the Egyptians 
themselves to be of the same race, and modern research has further 
shown that the Berbers and Tuaregs belong to the same physical 
and linguistic stocks as all the other Hamites Bejas, Afars, 
Agaos, Somals, Gallas, Masai who throughout all 
recorded time have occupied the eastern seabord 
from the equator to the Mediterranean, interrupted 
only by the Himyaritic Semite intruders in Abyssinia. The Afars, 
better known as Dankali (pi. Danakil), who hold 
the low-lying steppe between the Abyssinian escarp- 
ments and the Red Sea, show " not a trace of prognathism," and 

1 A. H. Sayce, Academy, April 14, 1894. 


are distinguished by "narrow straight nose, thin lips, small pointed 
chin not retreating, cheek bones not prominent, dark brown iris, 
pure white sclerotica, thick crisp hair, features South European 1 . 1 ' 
So the wide-spread Beja family Abdbdeh, Bishdri, Hadendawa, 
Homran, Beni-Amer and many others whose do- Bc - Aa 
main extends from that ofthe Afars northwards 
between the Nile and the Red Sea, into Upper Egypt, and who 
have been identified with the Macrobii of Herodotus, "tallest 
and finest of men" (in. 17). In any case they are physically a 
magnificent race, with well-shaped muscular frames, tall stature, 
"of European type, often very handsome, of a bronze, swarthy or 
light chocolate complexion, with long, crisp, but not woolly hair, 
generally falling in ringlets over the shoulders 2 ." 

Despite a perceptible strain of Negro blood, conspicuous espe- 
cially towards the ethnical borderlands, both the 
Gallas and their Somali cousins belong also funda- somais? &nd 
mentally to the same eastern Hamitic branch of 
the Caucasic division. Of all Hamitic peoples the Gallas, who 
call themselves Ilm'orma, "Sons of the Brave," are by far the 
most numerous, being estimated at from 7,000,000 to 8,000,000, 
spread over a territory of some 400,000 square miles, including 
the whole of South Ethiopia (Gallaland proper), besides large 
tracts in North Ethiopia (Abyssinia), and most of the little known 
region which extends through the Lake Rudolf (Samburu) de- 
pression to and beyond the Tana river. The typical Galias of 
KafTa and surrounding uplands are perhaps the finest people in 
all Africa 3 , tall, of shapely build, with high broad forehead, well- 

1 " Le fattezze sono europee del Mezzogiorno " (Fr. Scaramucci and E. H. 
Giglioli, Notizie sui Danakil, 1884, p. 5). At p. i, the features are said to be 
" Caucasian " despite a strain of Negro blood, as amongst all these eastern 

2 Linant Bey (Linant de Bellefonds), UEtbaye, pays habitt par les Bichariek, 
1868. The Bejas are the Bugavt the Axumite inscriptions, and the jSA^u/xm 
ofStrabo(i7, 53). 

3 "La race galla est la plus belle de l'Afrique...Les Gallas sont, en general, 
bien constitues. Us ont une haute taille, le front large et eleve", le nez aquilin, 
la bouche bien coupe*e, le teint cuivre plutdt que noir " (Rochet d'Hericourt, 
\er Voyage^ p. 174). So also Capt. Lugard : "a wonderfully handsome race, 
with high foreheads, brown skins, and soft wavy hair, quite different from the 
wool of the Bantus" (Proc. R. Geograph. Soc. 1892, p. 821). 



formed mouth, Roman nose, oval face, coppery or light chocolate 
colour, black kinky hair, often worn in " finger curls " or short 
ringlets round the head altogether noble representatives of the 
Caucasic family. In general the features are quite European, 
and even the complexion is no darker in some districts than that 
of the Andalusian peasantry 1 . TheoSomals also, whose domain 
comprises nearly the whole of the eastern horn of Africa, " are a 
very handsome race, of good physique, with excellent features 3 ." 


(East Hamitic Type.} 

By F. L. James they are " allied to the Caucasian type 3 ," but 
owing to secular interminglings with Negroes, Arabs, Afars, 
Abyssinians and other conterminous peoples, it is difficult to 
determine a general Somal type. The colour varies from light 
brown to black, and it is noteworthy that the darkest groups often 
present the most regular features 4 . 

Farther south and west the eastern Hamites are represented 

1 DrBeke, Jour. A\ Geograph. Soc. Xiv. p. 19. 
L> Commander F. G. Dundas, Geograph. Jour.> 1893, p. 211. 
<J The Unknown Horn of Africa, 1888, p. 7. 

4 "On dirait un beau sujet europeen dont la peau serait noire" (G. Revoil, 
Bull, de la Soc. de Geograph^ 1880, p. 259). 


by the Masai and blue-eyed Rendileh, and in the 
lake region by the Wa-Huma pastors, who, under w^-Huma!* 
diverse names (Watusi, Wahha, Wajiji, Warundi, 
Waruanda, &c.) are met scattered in small groups as far south as 
Lake Tanganyika. The Wa-Huma of Uganda, who are certainly of 
Galla descent, are described by Capt. F. D. Lugard as "tall, thin, 
and lithe, with high foreheads and most intelligent faces ; the eyes 
piercing, the features sharp, the nose often aquiline. In colour 
they vary, as do the Somals, some being very pale, others black. 
Some are remarkably handsome men. . . .They were much struck with 
the Somals [in camp], who, they said, must be of the same race 
as themselves 1 ." They hold themselves aloof from the surrounding 
Negroid populations, and despite the now prevailing dark shades, 
it is significant of their Hamitic origin that " the Waruanda call 
themselves white men, and deny all connection with the Bantu 
tribes 8 ." Intermediate between the Wa-Humas and the Gallas 
proper are the Masai, some of whom, such as the Ngaj, Molilian 
and other full-blood tribes, are " the most magnificently modelled 
men conceivable. ... In most cases the nose is well raised and 
straight, as good as any European's, though passing into the 
Negro type in the lower class, such as the Wa-Kwafi....The jaws 
are rarely prognathous, while the hair is a cross between the Eu- 
ropean and the Negro 8 ." Indeed an admixture of black blood is 
evident enough, despite the statement of Lieut, von Hohnel that 
" there is nothing of the Negro type in their appearance 4 ." 

The presence of this element is still more conspicuous in 
Abyssinia, where the blends between Negroes, Ethnlcal 
Hamites, and Semites are so multifarious and wide- Relations in 
spread that here nearly all the distinctive physical yss ma " 

1 The Rise of our East African Empire, i. p. 158. 

2 M. Lionel Decle, The Watusi, in Jour. Anthrop. fnst. May, 1894, p. 424. 
"The pure types," says this observer, "have long thin faces, with a long 
fine nose and a small mouth; their colour is of a rich brown without the 
violet black tints usually found in the Bantu races.... The hair does not grow in 
woolly patches of a dull colour, but is of a glossy black evenly spread all over 
the head... very like the hair of the Abyssinians.,.In fact they appear to me like 
a kind of connecting link between the Abyssinian and Bantu types " (/.). 

3 Joseph Thomson, Through Masailand, 1884, p. 427. 

4 Discovery of Lakes Rudolf and Stefani*, 1894, vol. I. p. 244. 


characters of the different races have lost their significance. 
" Neither the colour nor the hair are regarded as important 
ethnical tests, and the length of the heel alone [Negro "lark-heel"] 
is held to be an undoubted proof of Negro origin V To under- 
stand the present ethnical relations, it should be remembered that 
throughout the historic period Abyssinia has been the seat of 
powerful states, in which the dominant people have always been 
the Semitic Himya rites from south-west Arabia (Yemen). From 
the seaport of Adulis, founded by them on the coast below 
Massawa over 2000 years ago, their progress may be followed 
along the sites of the ancient cities of Koloe, Ava and Axum, 
successive centres of their power during the first centuries of the 
new era. The indigenous populations, with whom they had to 
contend, were mainly Hamites, one large section of whom, the 
Agao 2 , are mentioned in the Relation of Cosmas (523 A.D.) as 
already at that time subject to the Axumite kings. But others 
long maintained their independence, and in the loth century were 
strong enough to expel the Meniiek dynasty from Axum, a 
turning-point in the history of Ethiopia. Then the seat of govern- 
ment was shifted from the northern province of Tigre to the 
central region of Amhara, and by the close of the iyth century all 
the Hamite aborigines as far south as Shoa appear to have been 
brought under the sway of the Negus Negust, " King of Kings," 
representative of the old Axumite empire. During the course of 
these events the ruling Semitic classes were being slowly merged 
with their Hamitic subjects in a common Abyssinian nationality, 
which has further been modified by a large infusion 

The present J . 

Abyssinian of negro blood due to the long-standing institution 
populations. ^ domestic slavery, as well as by contact with the 
Galla Hamites, who for over 300 years have been encroaching on 
the southern and central provinces from South Ethiopia. Thus 
the present inhabitants of Abyssinia proper form an extremely 
complex ethnical group, in which it is not always possible to dis- 
tinguish the constituent elements. The prevailing colour is a 

1 De Quatrefages, op. cit., n. p. 395. 

2 Cosmas writes 'A^aO, and the name has been identified with the Athagao 
of the Adulis Inscription. It survives in the name of the large province of 
Agaomtdir, " Agaoland," still mainly inhabited by these primitive Hamites. 


_____,_, s _ 

distinct brown, shading northwards to a light olive and even fair 
complexion, southwards to a deep chocolate and an almost sooty 
black. There are Abyssinians who may certainly be called black, 
and in whom the negro strain is revealed in the somewhat tumid 
lips, small nose broad at base, and frizzly black hair. But the 
majority may be described as a mixed Hamito-Semitic people, 
who beyond question belong fundamentally to the Caucasic 

Thus is established the substantial ethnical unity of the indi- 
genous Hamitic populations of North and North- , . 

JT r- Relations of 

east Africa, as well as their direct relationship with the Hamites to 
the prehistoric inhabitants of Europe. Recent re- the Semites - 
search tends further to show that these Hamites formed originally 
a single ethnical group with the Semites of south-west Asia, and 
philologists already speak of an organic connection between the 
Hamitic and Semitic linguistic families. Hamitic, of which there 
are three recognised groups Old Egyptian with Coptic; Berber, 
including the Kabyle of Algeria, Shluh of Marocco, and Tamashek 
of the Sahara ; Ethiopian, current in a great diversity of forms 
amongst the Gallas, Somali, Agaos, Afars and Bejas belongs to 
the inflecting order of speech, and presents numerous points of 
contact with Semitic. The resemblance, however, is rather in the 
identity of their morphological base, than in the coincidence 
of fully developed grammatical forms. The subject is fully dis- 
cussed by Dr Fritz Hommel l , who establishes a close relationship 
in their phonetics, lexicography and structure between Semitic 
and Old Egyptian, and thus inferentially between both families. 
The pronominal systems are certainly alike both in their roots and 
in the process of plural formation ; internal vowel change is also a 
common feature, though much more highly developed in Semitic 
than in Hamitic; both attach the pronominal elements in the 
same way to the persons in verbal inflection, and both mark the 
feminine both in noun and verb by the same letter /. In Berber 
this element is both prefixed and suffixed, as in akli, negro; 
taklit, negress. 

1 Der babylonische Uf 'sprung dcr dgypfischen Kultur, Munich, 1892 ; and 
in Beitrdgt zur A$syrio/ogie II. Heft _, 1892. 


This fundamental unity of speech points at fundamental racial 
unity in two groups occupying conterminous domains from all time, 
and otherwise closely resembling each other in their more salient 
physical characters. Support is thus lent to the views of those 
anthropologists who are disposed with Hommel to bring the 
Hamites from Asia, or with Prof. Jastrow, Dr Brinton 1 and others 
to find the cradle of the Semitic race in North Africa. The latter 
view will be held to be the more probable by those who regard 
Mauritania as the original home and centre of dis- 
persion, not only of the Hamites, but of the Caucasic 
division itself, of which the Semites form one of the 
'chief branches. Yet until comparatively recent times the Semitic 
domain was mainly restricted to the south-west corner of Asia, 
that is to say, the region comprised between the Iranian plateau 
and the Persian Gulf on the east, and the Red Sea and Mediter- 
ranean on the west, with no clearly defined limits towards the 
north. From this relatively narrow territory the Semites spread 
in prehistoric times to Abyssinia, and along the southern shores of 
the Mediterranean to and beyond the "Pillars of Hercules." 
Later the Arab Semites overran nearly the whole of North Africa, 
formed settlements along the East African seabord south to 
Sofala, and penetrated eastwards to Persia, Central Asia, India 
and Malaysia. Apart from the doubtful Hittites, there are five 
great historical groups: i. The Assyrians of Mesopotamia; 
2. The Arameans (Syro-Chaldeans) of Syria, parts 
of Palestine and the Lower Euphrates; 3. The 
Canaanites (Hebrews, Phoenicians, Carthaginians, 
and others) of Palestine and the Mauritanian seabord with 

1 At a meeting of the Philadelphia Oriental Club 1890, papers on The 
Cradle of the Semites were read by Dr D. G. Brinton and Prof. Jastrow, the 
former contending that the Semitic stock came originally from " those pic- 
turesque valleys of the Atlas, which look forth toward the Great Ocean 
and the setting sun." While agreeing generally with this view of a probable 
Semitic migration from Africa to Asia, Prof. Jastrow held that there is not 
sufficient evidence to determine the particular region of Africa whence the 
dispersion took place. In fact, as here advocated, the whole area from the 
Mediterranean to Sudan must be included, as the Sahara presented in post- 
pliocene times a favourable milieu for the evolution of the highest human 


settlements in Iberia, the Mediterranean islands, and Bahrein ; 
4. The Arabs of the greater part of the peninsula named from 
them; 5. The Himyarites of Arabia Felix (Yemen) and Abys- 
sinia. Of these groups all but a few Syro-Chaldeans, the Hebrews, 
and the Abyssinian Himyarites, have either disappeared, or else 
been assimilated in speech to the Arabs, who may be said to have 
absorbed nearly all the other members of the Semitic family, much 
in the same way that the Latins absorbed all the other members 
of the old Italic family (Oscans, Samnites, Sabines, Umbrians). 

The Semitic type, as best represented by the Assyrians of the 
ancient monuments, by the Jews and by the Arabs, Semitic phy- 
offers considerable diversity in the details but is es- sicai and mcn- 
sentially Caucasic in its main characters, being distin- 
guished by perfectly regular and expressive features, fine oval face 



(Semitic Type.} 

and brain-cap, large and often aquiline nose depressed at the root, 
small pointed chin, forehead straight but not high, black almond- 
shaped eyes, dolichocephalic head, glossy jet-black hair, full beard, 
skin pale white but easily bronzed by exposure, stature rather 
below the average European (5 ft. 4 or 5 inches). This type, 
which in the upper classes often assumes an almost ideal beauty 

394 . ETHNOLOGY. . [CHAP. 

fully on a level with the highest European standard, approaches 
nearest to the Hamitic, at least as represented by the Mauritanian 
Berbers, from which it differs chiefly in the more perfectly oval 
contour lines of face and head. Compared with the Aryan, the 
Semitic intellect may be described as less varied, but more intense, 
a contrast due perhaps to their monotonous and almost changeless 
environment of yellow sands and blue skies, with a flora and fauna 
limited to a few species, and these mainly confined to oases and 
steppes encircled by the desert and everywhere presenting the same 
uniform aspect. Hence to the Semites mankind is indebted for little 
philosophy and science, but for much sublime poetry associated 
with many profound conceptions of a moral order, resulting in the 
three great monotheistic religions the Jewish, Christian and 
Muhammadan. Expansion and progress are the dominant cha- 
racteristics of the Aryan, concentration and immutability of the 
Semitic intellect. 

This mental temperament finds its outward expression in the 
Semitic form of speech, which is distinguished above 
10 a ^ others for great stability and persistence; so 
much so that the various branches (Assyrian, Ara- 
maic, Hebrew, Arabic, Himyaritic) may be regarded as mere 
dialects of a long extinct Semitic mother-tongue. They differ less 
from each other Hebrew, for instance, from Syriac, or Assyrian 
from Arabic than do many members of the same branch in the 
Aryan family English from Gothic in the Teutonic, Hindi from 
Sanskrit in the Indie branch. "On comparing the Chaldean of 
the fragments of Esdras, representing the Aramaic of the 5th 
century B.C., with the Syriac still written in our day, scarcely any 
essential differences can be detected between texts composed at 
so long an interval. Between these two limits Aramaic may be 
said to have varied no more than the language of Cicero from 
that of Ennius V Semitic speech presents some most remarkable 
phonetic and structural features, such as the series of deep gut- 
turals (kh, hh, q, gh) unpronounceable by Europeans, and conse- 
quently of racial value; and the verbal roots, mainly triliteral, 
" moved " by vowels, but never changed in sound or sequence in 

1 Kenan, Histoire...dcs langtics shnitiqncs* 


any of the branches; thus from root g//=kill, Arab, qatala, Heb. 
qdtal &c., "he killed." The whole verbal process, based on 
endless modifications of these roots within the prescribed, limits, 
is without analogy in any other linguistic system, and presents 
structural phenomena which have hitherto defied all attempts at 
analysis. From the triliteral root were developed, chiefly by 
internal vowel change and prefixed servile letters (^, /, , s\ as 
many as 15 thematic forms (intensives, reciprocals, causatives, 
reflexives, iteratives &c.), in the Semitic mother-tongue, of which 
12 or 13 are preserved in Himyaritic, n in Arabic, 5 in Hebrew 
and more or less in the other branches. Thus Arab, qatala, he 
killed; qutala, he was killed; quttala, he was utterly killed; 
qdtala, haqtala, taqatala, hinqatala, histaqtala &c., each with its 
personal endings, gender, participles, but two tenses only, the 
complete and incomplete. Peculiar to the Arabic branch is 
another striking feature, the so-called " broken plurals," on which, 
being really singular collectives, secondary plurals may be built. 
There are over thirty typical forms, such as jauhar, a gem, 
jawdhir, jewellery; amir, prince, umard, the nobility; kdfir, un- 
believer, kuffdr, the infidel ; qarib, a relation, aqribd, kindred, &c. 
Analogous forms survive in the cognate Himyaritic (Geez of 
Abyssinia) ; but the principle on which they have been developed 
has not been traced to any other member of the Semitic family. 

It will be noticed that in the Caucasic Family Tree no room 
has been found either for " Aryans/' or for the The A an- 
equivalent expressions " Indo-Europeans " or " Indo- speaking 
Germans 1 ." These are essentially linguistic, not peopes< 
racial terms, and the failure to distinguish between the groups of 
Aryan languages and the peoples of Aryan speech may be said to 

1 Strictly speaking Aryan, associated with the Airy ana Vaega of Hindu 
and the Ee*ryene Veejo of Persian traditions, is applicable only to the Indo- 
Iranian branch ; but its convenient extension to the whole group is too long 
established to be now set aside, especially as the alternative expressions Indo- 
European and Indo-Gcrmanic are themselves equally defective. They are 
purely geographical terms, and are far from covering the whole field, leaving 
out Irania and those other parts of Central and West Asia where Aryan- 
speaking peoples are indigenous. 



lie at the base of the prevailing confusion regarding the ethnical 
relations in Europe, Irania, and India. It is not denied that, at 
some remote prehistoric epoch, there was evolved in some Eurasian 
region a community of Caucasic type and of primitive Aryan 
speech, to which might properly be applied the expressions Aryan 
or Indo-European. But for ages these expressions have lost their 
full value, and Aryan race and Aryan speech have long ceased to 
be convertible terms 1 . The language has persisted under diverse 
forms down to the present day, and indeed is now the dominant 
speech of the world. But the primitive community, with whom it 
originated, has disappeared as a distinct ethnical group, dispersed 
so to say amid the innumerable populations on whom it imposed 
one form or another of the Aryan mother-tongue. This process 
could have been effected only by migrations and actual contact, 
if not conquest, resulting in the absorption of the intruders and 
the survival of their language amongst the masses reduced or 
influenced by them. Thus* it will be correct to say that an Aryan 
strain permeates all or most of the groups now speaking Aryan 
tongues, but not that these groups are themselves of Aryan stock. 
For it is to be remembered that, when the primitive Aryan man 
was first slowly evolved, the habitable globe was already fairly 
peopled by the diverse races, which, as we have seen, had established 
themselves in neolithic and even palaeolithic times in the regions 
stretching from India to the shores of the Atlantic. 
fmcte?of h the Hence the Aryan migrations cannot be conceived 
Aryan migra- as successive swarms going forth from some pri- 
meval Aryan cradleland, and for the first time 
peopling a great part of the northern hemisphere. Had these 
been the relations, the unity of the race, as well as of the language, 
would necessarily have been preserved. But the ground being 

1 M. de Nadaillac asks, who are the Helvetians? the Gauls? the Kelts? 
the Scythians and Cimmerians? And in reply to those who make them 
members of "la grande famille aryenne," he remarks that "les Aryas pas 
plus que les Semites ne sont un peuple ou une race ; ils forment une agglo- 
meration d'hommes unis par des rapports linguistiques " (Rev. des Questions 
Scientifiques-t Oct. 1894, p. 514). He should have said far less than the Semites, 
who do present, amid considerable diversity, a certain physical uniformity 
sufficient to constitute them a tolerably well-defined ethnical group. 


everywhere, or almost everywhere, already occupied, the wandering 
Aryan tribes could not fail to form fresh ethnical groups with the 
indigenous inhabitants, and thus sacrifice their own racial purity. 
Thus it happens that throughout the historic period various 
branches of the Aryan stock language have been, and still are, 
spoken by almost every ^ariety of the Caucasic division, by tall 
and short brachycephali, long-headed and round-headed Teutons 1 , 
all called " Germans " because of their German idiom ; by " Kelts " 
of so many types that the word has long ceased to have any ethnical 


significance; and by Armenians and Afghans often resembling 
Semites far more than ordinary " Aryans." We now see that it 
could not be otherwise, because, as explained in Chap. IX, the 
contact of two races speaking two distinct languages ultimately 

1 The typical German skull, as seen in the prehistoric graves, was highly 
dolichocephalic (mean index 71*3); yet at present brachycephaly increases 
continuously in the direction from north to south, so that in Bavaria it is almost 
universal. "Les Allemands du Sud sont essentiellement brachycephales. 
En Baviere, entre autres, Ranke a trouve que dans la plaine le nombre des 
individus presentant ce caractere est de 79 pour 100 ; sur les contreforts des 
montagnes la proportion monte a 83 pour 100 ; dans la montagne elle s'eleve 
a 90 pour 100" (De Quatrefages, op. cit. n. p. 490). 


results in the fusion of the races, but not of the languages, one 
of which must eventually prevail to the exclusion of its rival. 

What occurred generally during the early Aryan migrations 
may be illustrated by what occurred in Britain 

bythe rS?. d after the wi ^drawal of the Romans and the arrival 
tonic invasion o f the Angles, Saxons, Frisians and other allied 
Teutonic tribes. The old idea that these invaders 
made a tabula rasa of the land, repeopling it with their own stock, 
is no longer seriously entertained by any one. We now know, on 
the contrary, that the Teutons merged everywhere in diverse pro- 
portions with the Romano-Britons. Dr John Beddoe, who has 
devoted his whole life to these researches, finds interspersed 
amongst the Teutons numerous traces not only of the so-called 
" Kelts," related to Csesar's Belgje, and distinguished by rather 
broad head, slightly receding forehead, arched nose, prominent 
cheek-bones, long oval face, thin lips, pointed and projecting cjiin, 
light hair and eyes, and tall stature, averaging 5 feet 9 inches 
but also of still more primitive peoples, neolithic " Ibero-Berbers " 
with long narrow head, dark complexion, flat narrow and square 
forehead, prominent mouth and cheek-bones, concave or straight 
nose, light or dark grey eyes, very dark and often curly hair, and 
short stature ; and even a " Turanian " or Mongoloid element, with 
oblique eyes and brows, concave or flat nose, straight black or 
dark brown hair, broad cheek-bones and narrow chins 1 . Yet all 
these races were in a few generations so completely fused 
together in a common nationality of Teutonic speech, that the 
greater part of Britain might be supposed to have been originally 
settled by the intruders from north Germany in the 5th century. 
So it was in India, Irania, Sarmatia and other parts of the 
the P resent Indo-European domain, where small bands 
Hindu inva- of Aryan speech imposed their language and culture 
sion of India. Qn ^ surrouri( j m g population s ? which have thus 
come to be regarded as of Aryan descent. In India Dr Gustav 

1 The Races of Britain, 1885, passim ; and two papers Sur VHistoire de 
TIndice Cephalique dans les lies Britanniques, contributed to L? Anthropologie, 
1894, Vol. V. Nos. 5 and 6. This authority thinks that, even including the 
later Scandinavian arrivals, the Teutonic element amounts to not more than 
about one half in the greater part of England. 




Oppert's investigations, spread over many years, tend to show 
that the Aryan invaders never were numerous, and that their 
influence on the aborigines was more social and religious than 
ethnical. Thanks to their higher culture and superior mental 
endowments, they imposed their religion on the masses every- 
where throughout the peninsula, and their Aryan speech (Sanskrit) 
on most of the populations in the Indo-Gangetic regions 1 . At the 
census of 1891 as many as 195 millions were returned as of Neo- 
Sanskritic speech, of whom probably not five per cent, were full- 


(High- Caste Hindu Type.} 

blood Hindus of the higher castes. Even the haughty Rajputs, 
formerly of the Kshatria (military) caste, have long lost their racial 
purity, and are now largely intermingled with Bhils and other primi- 
tive non-Aryan races. The same process has been in progress for 
many ages on the Sarmatian plains, where the Scythian and other 
Mongoloid hordes have been gradually Aryanised by peoples of 
Slav and Gothic speech. The Gothic language, which still sur- 
vived in the Crimea down to the i6th century 2 , is now extinct, 

1 On the Original Inhabitants of Bharatavarsa or India, 1894, passim. 

2 H. Bradley, The Goths, 1888, p. 363. A list of words is here given, 
taken down by a Belgian traveller in 1562, from which it is evident that the 


and nearly all these populations speak Great Russian and Little 
Russian dialects. Hence they pass for Aryan Slavs, just as the 


short, dark, brachycephalic peoples of Auvergne and Savoy pass 
for Aryan Kelts. 

From these considerations it would seem that somewhat undue 

importance has been attached to the quest of the 
crldfeit^ an Aryan cradleland, which in recent years has been 

prosecuted with so much zeal, in the belief that 
here would be found the original home of the multitudinous 

language of Wulfila (Ulphilas) was still current in Taurida at that time. Such 
are mine, mycha, wichtgata, ies Goth, mtna, meki, hweitata, is (moon, sword, 
white, he). Of the Sarmatae (Sauromatae) nothing positive can be asserted, 
though it would appear that those known to Tacitus (Germania, ch. i) were not 
of Mongolo-Tatar speech. Probably the bulk of the nation was originally of 
Mongol stock, but had at that time already been brought under Aryan (Slav?) 
influences. It is noteworthy that in their territory (South Russia) many recent 
ethnologists are disposed to place the primeval home of the Aryan race. But 
the question is beset with so many difficulties that those only who do not know 
' venture to speak confidently. Thus the primitive Slav type appears to have 
been decidedly dolichocephalic, t whereas the present Slavs are mainly brachy- 
cephalic (Ch. de Ujfalvy, Le Berceau des Aryas, etc., p. 25). 


populations now speaking Aryan languages. Nevertheless, such 
an inquiry can never be devoid of interest, as bearing on 
the centre of evolution of a gifted prehistoric people who, 
more than any other, may be supposed by their very dispersion to 
have leavened the rude prehistoric masses, thus raising a great 
part of humanity to a higher social plane. But it is not to be 
imagined that the primitive Aryan groups stood themselves on 
a very high level, when they began to break up and spread 
abroad amid the surrounding populations. Their assumed superi- 
ority would seem to have been rather potential than 
actually established, and the organic elements of Aryan"cuiture. 
their speech show that, before the dispersion, they 
were a rude pastoral people, possessing cattle, sheep, goats and 
the watch-dog, but with scarcely a rudimentary knowledge of agri- 
culture. They were half troglodytes, dwelling in winter in holes 
dug in the ground and roofed with turf, in summer either in 
round 'huts made of poles with interwoven branches, or in lum- 
bering waggons with wheels and axle chipped and charred from 
a single stem. Originally they wore undressed skins, giving place 
later to garments roughly woven of wool and flax. They also 
made rude earthenware, but lived in the polished stone age with 
no knowledge of the metals, except perhaps copper, used more for 
ornaments than for weapons. The bride was captured or purchased, 
and the family was based on polygamous and patriarchal insti- 
tutions ; nor can there be any doubt that " ancient Aryan custom 
ordained that the wife should die with her husband," while " the 
custom of putting a violent end to the aged and infirm survived even 
into historic times 1 ." It appears also that these primitive pastors 
dwelt in an open region with a continental climate, that is to say, 
severe winters and hot summers, so that they recognised but two 
or three seasons 2 , reckoning the years as "winters " divided into 
"moons" and "nights," not months and days, and making no 

1 F. B. Jevons, Prehistoric Antiquities of the Aryan Peoples, being the 
14 Sprachvergleichung und Urgeschichte of Dr O. Schrader," 2nd revised ed., 
1 889 90. 

2 Schrader says "two or three" ; but van den Gheyn shows that "on a 
la preuve manifesto de 1'existence de trois saisons, le printemps, 1'^te, Thiver, 
chez les Aryas primitifs " (IS Origins eitroplenne des Aryas t 1885, p. 1 1). 

K. 26 


attempt to harmonize solar and lunar time. By the application 
of this "linguistic palaeontology," as Pictet has called it, it is 
further concluded that the Aryan cradleland was in the nature of 
a steppe with a marked absence of mountains and continuous 
forests, but traversed by numerous broad and shallow watercourses, 
and with a poor flora distinguished by such hardy growths as the 
birch, poplar, willow, reeds and rushed. 

Such are the general deductions drawn by Dr Schrader from a 
careful study of the common elements of primitive 
Aryan speech, and it is because these conditions 
appear to prevail to a greater extent in the South 
Russian steppes than elsewhere that, after some hesitation, this 
authority finally concludes " that the scene of the most ancient 
period of Indo-European development, the original home of our 
race, is to be looked for " in that region (/#.). This conclusion, 
however, has not been so generally accepted as is commonly sup- 
posed, and it has been rejected not only by most French anthro- 
pologists, but also by Ch. de Ujfalvy 1 , Briinnhoffer 2 , Orterer 3 , von 
Roth of Tubingen 4 , van den Gheyn fi and others, who still hold by 
the Asiatic view first attacked by Latham towards the middle of 
the i Qth century. 

But a glance at the map of Eurasia, with a consideration of 

Conflicting ^ e c ^matic conditions prevalent in the northern 

views regard- hemisphere in prehistoric times, may help to recon- 

ingthe Aryan , . , . ^ 

Cradleland re- cile both theories. In the space stretching from the 
conciied. Urals to the Caucasus there are no natural barriers 

between the two Continents, while the Urals themselves are much 
too low and too gently inclined to offer any serious impediment to 
the migrations of primitive man, who was free to roam everywhere 
over the Aralo-Caspian depression and the Sarmatian plains 
from the Turkistan highlands westwards to the Carpathians. 
At present moisture decreases continuously eastwards throughout 

1 Lc Berceau des Aryas cTaprls des Ouvrages Rtcents* 1884. 

2 Ueber den Ursitz der Indogermanen^ 1884. 

8 Literarische Rundschau^ 1884, No. 9, pp. 267 et seq. 
4 Zeitschrift der D. M. G. xxxvin. p. 138. 

r> Les Migrations des Aryas, 1882; UOrigine europtenne des Aryas, 1885, 
and other writings. 


the whole of this steppe region, and the process of desiccation has 
been going on for ages generally throughout Central Asia. Thus 
the Zerafshan, the Dehas (Baikh), the Mtirghab, the Tejend and 
many other streams, which formerly reached the Oxus or the 
Caspian, now run out in the Kizil-Kum, the Kara-Kum and 
other sandy wastes of the Turkistan depression. But in neolithic 
times Turkistan was as suitable for human habitation as the 
South Russian steppes still are, and the long-abandoned pre- 
historic highway leading from Hyrcania to Baktriana is strewn 
in some places with numerous ruins now mostly buried under the 
surging sands of the wilderness. "The local traditions, historical 
records, and the ruins of numerous cities leave no doubt that the 
country was formerly far more densely peopled. The inhabitants 
have disappeared with the running waters ; the powerful empires 
of the Oxus and Sogdiana basins have vanished ; the great centres 
of Eastern civilisation have become eclipsed; many cultured 
peoples have reverted to barbarism and the nomad has triumphed 
over the agricultural state V 

In other respects there was nothing to choose between the 
eastern and western sections of the Eurasian plains, The Eura . 
both of which equally presented the climatic, bo- sian steppe 
tanical and other natural conditions reflected in the C primitive 
the common elements of Aryan speech 2 . It will Ar y an e rou P s - 
therefore be more reasonable to place the Aryan cradleland in 
this Eurasian steppe region generally than restrict it either to the 
European or to the Asiatic sections, separated as these are by 
purely conventional limits. So difficult is it to draw any hard and 
fast line between the two zones, which are essentially one from the 
physiographical standpoint, that the present Russian government 
of Orenburg actually comprises both slopes of the southern Urals, 
that is, includes parts of Europe and Asia in the same admini- 
strative province. 

1 Reclus, English eel., VI. p. 162. 

2 " On fait beau coup valoir pour la provenance europeenne des Aryas les 
exigences du climat, de la faune et de la flore, reVelees par la paleontoiogie 
linguistique dont les donnees rclament une contree relativement froide. Faut- 
il sans cesse affirmer que ces conditions sont re'alise'es en Asie centrale? M (Van 
den Gheyn, L'Origine europeenne des Aryas t pp. 42 3). 

26 2 




Thus may also be explained the apparently contradictory state- 
ments that the primitive Aryan communities were long in close 
contact with the Semitic peoples, while on the other hand their 
primitive culture was apparently the same as that which prevailed 
both in the early lacustrine settlements of Switzerland 1 and in the 
so-called terramare, or prehistoric stations of North Italy. As 


(Irano-Semitic Type.) 

pastoral nomads these Aryan groups needed a vast space for the 
support of the numerous herds on which their existence depended. 
Thus while some roamed westwards and gradually penetrated up 

1 In a paper contributed to the Revue des Questions Scientifiques for 
October, 1894, on Les Populations Laciistres de V Europe ^ M. de Nadaillac is 
disposed to associate these settlements with the first Aryan wanderings in 
neolithic times, and to give them an Asiatic origin. At the same time he 
admits that no distinct traces of Asiatic art, except perhaps the somewhat rare 
objects made of nephrite, have been discovered in the dttris, though he is 
inclined to think that the jade objects are also more probably of Asiatic than 
of European origin, adding, "si la nephrite et la chloromelanite ont ete' 
importees d'Asie, pourquoi n'en serait-il de meme pour les jadeites?" (p. 500). 
But far too much importance is attached to these questions, which could never 
prove anything more than commercial intercourse, such as is known to have 
already been established in remote prehistoric times between the Black Sea and 
the amber-yielding shores of the Baltic. 


the Danube to and beyond the Swiss valleys, others advanced 
from the Turkistan steppe to the Iranian plateau and the head 
waters of the Euphrates, where they found themselves contermi- 
nous with the territory of the Assyrian Semites. Hence the 
distinctive hooked nose and other Semitic features still prevalent 
amongst the Armenian, Circassian and Iranian populations through- 
out the whole region between the Euphrates and Indus basins. 

But when the cradleland is found the primitive Aryan group 
itself still eludes our grasp. As well seek in the raised dough the 
leaven of fermentation, as try to determine a primitive Aryan type. 
From the foregoing remarks it must be obvious 

. , , , . , The primi- 

that those described by ethnologists as many as tive Aryan type 
six are spoken of are types, not of the original Determine 
Aryan groups, but of the present Aryan-speaking 
peoples, and Virchow's challenge remains unanswered : " Who 
therefore will furnish the proof that the primitive Aryans were all 
doiichocephalous and had blue eyes, blond hair and a white 
complexion 1 ? " It cannot be too strongly insisted upon, not 
only that the world was peopled before the Aryan dispersion, 
but also that tall, fair, long-headed peoples, such as are usually 
regarded as typical Aryans, had already been evolved in North 
Africa, and had thence spread over West Europe and Scandinavia 
while the Aryan nomads were still tending their flocks and herds 
on the Eurasian steppe lands. There were also other non-Aryan 
peoples in Europe long before that region was reached by the 
Indo-Germanic hordes. Such especially was that short, round- 
headed dark race, which has been called both " Kelt " and 
" Lapp V an d which is still represented by the Low Bretons, the 

1 Die Urbevolkerung Europe? s % p. 33. 

a In one place de Quatrefages treats the round-headed populations of Savoy 
(Aix and Chambery districts) as "tout au moins extr&nement voisines des 
Lapons" (op. cit. p. 455) ; in another these Savoyards "touchent de plus ou 
moins pres a la race celtique," and are identified with the highland Galchas 
described by de Ujfalvy and Topinard. The Galcha skull measured by 
Topinard is said to present, (( non plus de simples ressemblances, mais une 
identit^ a bien peu pres complete avec les cranes les mieux caracterise's de 
Savoyards" (ib. p. 489). Topinard's language is very strong : "La reproduc- 
tion frappante du type Savoyard que nous regardons aujourd'hui comme une 
expression de Pancien type celtique, plus parfaite encore que le type auvergnat 


Auvergnats, the Savoyards, the Croatians and, as shown by de 
Ujfalvy, by the Galcha Highlanders of the Hindu- Kitsh and 
Turkistan uplands. Thus wherever they presented themselves 
the Aryan tribes could only play the part of intruders, intermingling 
with the aborigines, imposing on them their speech and culture, 
and modifying in various degrees their physical type. 

Nevertheless, all things considered, it seems probable enough 

that the typical Aryans belonged rather to the 

But probably X anthochroid than to the melanochroid branch of 


the Caucasic division. This may be inferred from 


(Iranian Type.) 

the distinct blond strain, which is found permeating most Aryan - 
speaking peoples in varying proportions, and which seems best 
explained by the assumption of an Aryan element grafted on those 

ou le type bas-breton" (Rev. cfAnthrop. Oct. 1878, p. 706). And thus the 
domain of the Keltic race is extended to the heart of Asia, while the whole of 
Europe is represented by Dr R. Cruel as occupied by "Turanian" peoples of 
Ural-Altaic speech before the arrival of the Aryans (Die Sprachen itnd 
Volker Europds vor der arischen Einwanderung, Detmold, 1883, passim). 


aborigines. Prof. G. de Lapouge aptly remarks that "no people 
amongst whom the fair long-headed type prevails makes use of 
non-Aryan languages and institutions, whereas the peoples where 
this type is not dominant make partial use of languages and 
institutions other than Aryan ; they have done so within a recent 
historical epoch (part of Pfissia and Germany), or appear to have 
done so in ancient times (Gaul, Spain). 1 " The reference here is 
to the Esthonian, Livonian and Krirland Finns of the Baltic 
provinces, now nearly extinct ; to the numerous groups of Volga 
and other eastern Finns not yet absorbed by the surrounding 


(Iranian Type.) 

Slav populations ; and in the west to the Iberi, Aquitani (?) and 
others of non- Aryan speech, now represented only by the Pyrenean 
Basques. Amongst all these the assumed Aryan element is 
perhaps less pronounced than amongst those more illustrious 
historical groups which are commonly regarded as full-blood or 
typical Aryans. Such are in the East the early Persians and the 

1 L'Origine des Aryens, Science, Aug. 4, 1893, p. 65. 


Hindus, who occupied the Iranian tableland and the Indus basin 
respectively some thousand years ago, and in the West the Hellenes, 
Teutons and others who under diverse names swarmed into 
South-east and Central Europe, if not in the neolithic and bronze 
periods, certainly in the first iron age, that is, the epoch repre- 
sented by the Hallstadt culture and ty the extensive sepulchral 


mounds (over 20,000) brought to light on the Glasinac (Glasinats) 
plateau near Sarayevo since the Austrian occupation of Bosnia *. 

To the close linguistic unity by which these widely- scattered 
groups are connected corresponds a certain ethnical unity, indi- 
cated especially by their common dolichocephaly and fair com- 
plexions, later obscured in many places by miscegenation with the 

1 A detailed account of this vast prehistoric necropolis is given by Herr 
Salomon Reinach in L* Anthropologie for September October, 1894, pp. 563 
et seq. 


aborigines of the regions severally occupied by them. It was 
above shown that the early Teutons were certainly long-headed, 
as were also the Hellenes 1 , while "the primitive Greek skull 
appears to present a very close resemblance to that of the high- 
caste Hindus, and this in its turn almost exactly reproduces that of 
the dolichocephalous Persians 2 ." 

It is difficult to resist the conclusion that we have here, if 
anywhere, the nearest approach to the original Aryan type, which 


(Iranian Type.} 

would have thus resembled that of the Afro-European as repre- 
sented by the Mauritanian Berbers, by Mr Petrie's " new race " in 
Egypt, and by early neolithic man in West Europe. To account for 
this remarkable coincidence, it is only necessary to assume a 
twofold dispersion of the primitive Caucasic groups from North 

1 Proved by the extensive researches of Sig. Nicolucci and Dr Hamy. The 
last mentioned describes a Greek skull of the loth Century B.C., now in the 
Paris Anthropological Collection; but "en Grece comme ailleurs, le type 
primitif a etc* altere par le croisement " (De Quatrefages, op. tit. p. 494). 

2 De Quatrefages, op. cit. p. 494. 


Africa, one northwards to Iberia and West Europe, the other 
eastwards into the Semitic domain (South-west Asia) and thence 
northwards to the Eurasian steppes. Both of these movements 
must in any case be accepted to account for the relations between 
Berbers and West Europeans on the one hand, and on the other 
between the African Hamites and th$ Asiatic Semites, as above 
set forth. 

To sum up this difficult Aryan question, the Aryan peoples 
__ A must be regarded, not as a single ethnical stock, 

The Aryan 

problem sum- but as an amalgam of many Caucasic, and no doubt 
medup. some Mongolic elements, leavened by an original 

xanthochroid strain, and endowed with a certain racial uniformity 
by the immense preponderance of the Caucasic physical charac- 
ters, and by the general adoption of Aryan speech, traditions and 
institutions. The process of fusion, resulting in the historic Aryan 
peoples, had its beginning with the first contact of the migrating 
tribes with alien races after the dispersion from a common cradle- 
land, and this process has never ceased throughout historic times. 
It is now developing new and often profoundly modified Aryan- 
speaking groups in North America (Franco- Canadian 
pansionofthe half-breeds), throughout Spanish and Portuguese 

America (Mestizos of all varieties), in South Africa 
(Dutch-speaking Hottentot half-breeds), in Indo- 
China (Franco-Annamese), in North Russia and Siberia (Russo- 
Finns and Russo-Tatars) and elsewhere. But as a rule the 
Anglo-Saxon or British Aryans, who are by far the most numerous 
and widespread out of Europe, do not amalgamate with the 
aborigines. Hence Anglo-American, Anglo-African and Anglo- 
Australian half-castes are rare, and the modifications of the Aryan 
types undoubtedly going on in the " Greater Britain " beyond the 
seas are due, not to miscegenation witti lower races, but partly to 
the changed environment, partly (North America) to fusion with 
Germans, Scandinavians, and other fellow-Aryans. 

In the Aryan linguistic family, in which root and formative 

The A & elements are, so to say, chemically combined, the in- 

Linguistic fleeting principle receives its most perfect expression 

Family. ^^ ^ ^ ^ branches (whj[ch recent ^ 

search has raised from eight to ten by the addition of the Galchic 


group 1 and the removal of Armenian 2 from the Iranic connection) 
spring directly, but in various divergent lines, from a primitive 
Aryan stock language long extinct past recovery, and all attempts 
at the reconstruction of which have proved abortive. The 
divergent lines represent each a distinct branch of the mother- 
tongue, and the divergence began at such a remote epoch that 
the mother-tongues of the several branches themselves are also 
irrecoverably lost. Not only so, but the earliest known forms of 
these different groups are already so profoundly differentiated from 
each other that their common relationship alone can be demon- 
strated; the order of their divergence from the parent stem, or 
from some now lost intermediate stems, remaining more or less 
conjectural. Each group comprises two or more subdivisions, 
which again throw off numerous branchlets, the whole forming an 
exceedingly complex system, which will be best understood by the 
subjoined : 

1 The researches of Ch. de Ujfalvy (Bui. de la Soc. de Geogr. June, 1878, and 
numerous other memoirs), of Major Biddulph (The Tribes of the Hindoo- Koosh), 
Robert Shaw (On the Galtchah Languages in. Jour. As. Soc. Bengal, XLV., 1876, 
and XLVI., 1877), Prof. Tomaschek (Die Pamir-Dialekte) and Prof. W. Geiger 
(Ostiranische Kultur im Altertuni) have placed beyond doubt the existence of 
numerous primitive Aryan languages on both slopes of the Hindu-Kush and on 
the western escarpments of the Pamir, to which de Ujfalvy has given the 
collective name of Galcha^ and which hold an independent position somewhat 
intermediate between Baktrian (Zend) and Sanskrit. " II est certain que les 
idiomes de 1'Asie centrale ont mis sur la piste de plusieurs formes intermediaires 
qui manquaient pour renouer la chalne parfois interrompue qui relie le Sanscrit 
au bactrien" (J 1 . van den Gheyn, Les Langues de VAsie Centrale, Leyden, 
1884, p. 27)- 

2 At a meeting of the Philological Society, May 13, 1892, the late MrG. A. 
Schrumpf read a paper on "The Place and Importance of Armenian in 
Comparative Philology," confirming the view of Prof. Hubschmann that 
the Krapar or old literary language of Armenia forms with its modern repre- 
sentatives an independent group in the Aryan family, distinct from Iranic with 
which it has hitherto been connected, and showing certain features intermediate 
between Iranic and Slavo-Lithuanic. It thus serves to bridge over the gap 
between the Asiatic and European divisions, and as Dr Fr. Miiller suggests, its 
nearest congeners may have been the Thrakian (and Phrygian) formerly current 
on both sides of the Bosphorus. Others however, with Karl Blind, affiliate 
Thrakian to the Teutonic branch, a view for which there is much to be 







Sanskrit \ The Prakrits 
(Vedic) > (Vulgar - 

Later / Sanskrit) 
.Sanskrit ' 


lc 2 

( KashmM 
kOriya; Assam! 




Westera| old p ersiallj p ah i4 vi> Neo-Persian, Kurdish, Baluchi 

/ Pamir ) Shignani, Iskashami, Wakhi, Sanglichi, Yagnobi, 
1 Group \ Minghani, Yidghah 

JHindu-KushJGilgit, Astor, Torwalak, Gowro, Bushkarik, 
V Group \ Narisati, Khowar, Bushgali 

( Old Armenian, Modern Armenian, Ossetian, Thrakian (?), 
( Phrygian (?) 

Old Illyric, Albanian: Tosk, 

^Eolian, Dorian, Ionian, Attic, Byzantine, Romaic 

V Group 

/ Oscan \ 

ITALIC J Sabine [ (Extinct) 
BRANCH jUmbrianJ 

\Latin, Vulgar Latin, Neo-Latin 

Langue d'Oc 
Langue d'Oil 

(Standard French) 




j Gsedhelic : Irish, Gaelic, Manx 

( Kymric : Welsh, Cornish, Low Breton 

Lithuanian, Lettic, Pruczi (Old Prussian) 

Eastern) Old Slavonic, Great Russian, Little Russian, Servo- 
Group j Croatian, Slovenian, Bulgarian 
Western ) Tsekh (Bohemian), Polish, Polabish, Lusatian, 
. Group { Slovak 

(Low German) Gothic, Frisic, Dutch, Continental Saxon, Anglo- 
Group ) Saxon, English, Lowland Scotch 
Group{ Old Norse ' Icelandic Norwegian, Danish, Swedish 
High German) Old, Middle and High German, Rhenish, 
Group \ Thuringian, Swiss, Suabian 

The profound disintegration which is shown in this Table, and 
which is immeasurably greater than in the Semitic 
tionofprimi- family, is mainly due to the spread of Aryan speech, 
amongst non-Aryan peoples, by whom its phonetic 
system and grammatical structure were diversely 



modified. But apart from these potent outward influences, all the 
Aryan tongues have, throughout their historic life, betrayed an 
inner tendency to break up the highly developed inflectional forms 
of the early languages, such as Sanskrit, Baktrian, Greek and Latin, 
and thus continue their natural evolution in the direction from 
synthesis towards analysis. Thus the Romance (Neo-Latin) 
gradually rejected all case* endings and passive verbal forms, and 
the Latin amabor, for instance, is now expressed by three words 
in Italian and French : io sarb amato \ je serai aime. It would 
require four in English (/ shall be loved), and in this respect 
English is the most highly developed, that is, the most analytical, 
of all Aryan languages, having retained scarcely a dozen of 
the many hundred inflections characteristic of primitive Aryan 

The Teutonic group, of which English is now the chief 
member, the Weltsprache as the Germans call it, _. _ ^ . 

* e I ne l eutonic 

presents some remarkable phonetic features which phonetic sys- 
give it an unique place in the Aryan family. In em * 
this group the organic Aryan mutes undergo two distinct series of 
permutations, in accordance with the so-called law of Lautver- 
schiebung ("sound-shifting") discovered by Rask, developed by 
Grimm and completed by Verner. The first series of shifts took 
place in prehistoric times, and is found already fully carried out in 
Gothic, the oldest known member of the group. In this process 
the surds or voiceless stops /, , / first become everywhere the 
voiceless spirants f, h, th ; then these spirants, when medial and 
in association with sonants, become themselves the sonant or 
voiced stops , g, d, always in weak syllables, and also in strong 
syllables before the accent ; but when they follow the accent the 
second shift is arrested, and they remain voiceless spirants. The 
influence of the Aryan accent, first noted by Verner, is seen in such 
examples as Sanskrit dntara, Gothic dnthar, Anglo-Saxon and 
English bther for hither, with single shift only (/ to tK) because the 
accent precedes ; but Sansk. antdr, Goth, undar, A.-S. and Eng. 
under, with double shift (/ through th to d\ because the accent 
follows. The process extends in A.-S. and Norse to the organic 
voiceless spirant s, which similarly passes through z to r, as in 
Goth, dius (for diuz), Norse djr, A.-S. debr> Eng. deer. The 


second series of shifts is historical, no trace of it occurring in the 
Gothic of Ulphilas (4th century), or in any extant Teutonic forms 
(geographical or personal names &c.) before the yth century. Its 
later appearance is also shown by the fact that it never spread to 
the whole of the Teutonic domain, but is mainly confined to the 
South German highlands, where the process was continued spo- 
radically to about the beginning of the i 2th century. The South 
German dialects were thus constituted a distinct group under the 
name of Hoch-Deutsch (" High German ") in contradistinction to 
the Platt-Leutsch ("Flat" or "Lowland German") of the northern 
plains, which were unaffected by the process, and which conse- 
quently remain in their phonetics truer representatives of primitive 
Teutonic speech. The process itself is due to a general tendency 
to strengthen the mutes, so that the soft sonants (, g, d) become 
hard surds (/>, , /), while these become hard (voiceless) spirants 
(Pfvtf* h or ck ts written z). Thus the Catti of the Romans pass 
through such forms as Chatti, Hatti, Hazi, Hassi to the modern 
Hessians. But the rotation is arrested at the hard spirants/ /*, th 
of the prehistoric series (representing organic /, k, /), because 
these are incapable of further strengthening. Hence it is that the 
primitive Teutonic / and h persist in High German (Gr. mjW, 
Goth, hunds, Germ, hund, Eng. hound). Surd ///, however, passes 
through sonant th (dh) to d, and later further changes take place 
in the Hoch-Deutsch group, which thus becomes differentiated 
into Old (yth to nth century), Middle (i2th to I5th) and Modern 
High German. In general the dental are much more fully carried 
out than the labial and guttural shiftings, so that the primitive surd 
th (as in thin) passes through sonant th (as in then) to d in the 
Low as well as in the High German group, but not in A.-S. and 
English, which thus stand phonetically on the same high level as 
Gothic itself, that is, nearest to the organic Aryan speech. Hence 
it is that words like three (Goth, threis, A.-S. thred), thorn (Goth. 
thaurnus, A.-S. thorn) &c., appear both in Low and High German 
with initial d\ Dutch drie, doom; Ger. <//', dorn; all represent 
ing organic Aryan /, as in Sans, tri, Gr. rpcts &C. 1 

1 A. H. Keane, Teutonic Languages^ in Cassell's Storehouse of Information^ 
1894, p. 243. 


In the Caucasus the ethnical and linguistic relations present a 
marked contrast to those prevailing in all other 

Ethnical and 

parts of the Caucasic domain. Probably more stock linguistic rcia- 
languages are current amongst the few hundred 
thousand natives of this relatively small moun- 
tainous region, than amongst the myriads of other Caucasic 
peoples spread over both hemispheres. The highlanders them- 
selves belong to the melanochroid division, and some of the 
groups, such as the Georgians, Circassians, Kabards and Les- 
ghians, approach an almost ideal standard of physical beauty. 
But considerable diversity prevails, and the Pshavs, Svanitians 
and others confined to less favoured districts, have coarse, almost 
repulsive features and ungainly figures. We have seen (p. 93) that 
the Caucasus was already occupied by palaeolithic man, although 
perhaps not to a great extent. But in the polished stone age the 
whole region appears to have been thickly inhabited, as shown 
by the numerous dolmens, lacustrine stations, and other remains 
of a culture closely analogous to that of neolithic man in Europe. 
According to the researches of M. Chantre these prehistoric 
peoples were long-headed, and in other respects resembled the 
Iranians and the high-caste Hindus. But, since then, great inter- 
minglings have taken place, with the result that the physical 
characters have been gradually modified in the direction of the 
brachycephalous dark type 1 . 

From this it would almost seem as if Caucasia was originally 
occupied by primitive Aryan tribes of the xanthochroid type, 
although of the numerous distinct languages now spoken in these 
uplands one only, the Ossetian, is a member of the Aryan lin- 
guistic family. All the others belong, not to the inflecting, but to 
the agglutinating order of speech, without however showing any 
clear relationship with the Uralo-Altaic or with any other linguistic 
family. They are usually grouped in three main divisions : the 
Southern^ comprising the Georgians, Imeritians, Mingrelians, Svani- 

1 " M. Chantre a mis en se*rie dix-sept indices moyens pris sur autant 
de populations anciennes et modernes; et on voit Tindice grandir d'ige 
en age, depuis 71-55 (Samthavro, premier &ge du fer) jusqu'a 86*48 (Ossettes 
de Koban modernes)" (De Quatrefages, op. at. p. 475). 



tians, Khevsurs, Pshavs and Lazes, all speaking 
distinct branches of a common stock language, sup- 
posed by Sayce to be the "Vannic," that is, the 
language of the Cuneiform Inscriptions of the Lake 
Van district. The Western, comprising the now dis- 
persed Circassians and Abkhasians besides the Kabards, Shapsukhs 
and others, also speaking languages believed to be derived from a 

Main divi- 
sions of the 
peoples and 
languages of 


(Melanochroid Type.} 

common source. The Eastern, comprising all the Daghestani 
peoples, Chechenzes, Lesghians, Avars, Galgai, Ingushes, Kishi, 
Tushi, Karabulaks, Kurini, Kubachi, Duodez, Ude, Dido, Dargo, 
Andi and many others, whose various idioms have hitherto resisted 
all attempts of the philologists to reduce them to a common stock 
language. Some may be grouped in a single family; but others, 
and especially the Ude, Kubachi, Andi and Dargo, must for the 
present be regarded as so many stock languages. In the South 
Daghestani tongues agglutination has reached such a high develop- 
ment that General P. V. Uslar, the first authority on this subject, 


calls them inflecting, which confirms the views advocated in Chap. 
IX. regarding the evolution of the higher from the lower orders of 
speech '. 

The greater portion of South India (the Deccan) is occupied 
by an indigenous people numbering over 50 millions, 
collectively known as DraJidas, and speaking dia- linguistic rcia- 

lects of an agglutinating stock language. The chief 
members of this group, whose ethnical position it is 
difficult to determine, are the Telingas (Telugus) of th* Northern 
Circars and part of the Nizam's territory; the Tamils of the 
Karnatic, South Travancore and North Ceylon ; the Kanarese of 
Mysore, the southern districts of the Bombay Presidency, and of 
Kanara on the Malabar Coast ; the Malaydlim, on the same coast 
south of the Kanarese ; the Kodagu of Kurg, west of Mysore ; the 
Ordons and Rajmahdli of Chota Nagpdr; the Gonds of Gond- 
wana, t Vindhya Hills. Although they preceded the Aryan -speaking 
Hindus, the Dravidas are not the true aborigines of the Deccan, 
for they were themselves preceded by dark peoples, probably of 
aberrant Negrito type (Chap. XI.). They are usually regarded as a 
Mongoloid people, who entered India from the north-west, leaving 
on the route the Brahuis of Baluchistan, whose language shows some 
remote resemblance to Dravidian. But at present the type cannot 
be called Mongolic ; it scarcely differs from the average Hindu, 
except in some districts, where it has been somewhat modified by 
contact with the Kolarians and dark aborigines. Hence they are 
grouped with the northern Hindus by Peschel, who remarks that 
" their most noticeable feature is their long black hair, neither tufted 
nor straight, but crimped or curly. This clearly distinguishes them 
from the Mongoloid natives, as does the fact that the hair of their 
beard and bodies grows profusely.... The inhabitants of India form 
at present but a single race, and the separation of the populations 
resident between the Himalayas and the Vindhya Mountains from 
the Dravidas of the Deccan is based solely on the fact that the 

1 Uslar's Memoirs on the " Caucasian Microcosm," as he calls it, are 
dispersed amongst the Bulletins of the Petersburg Imperial Academy of 
Sciences, and in the publications of the Imperial Geographical Society. But 
a useful summary Sur I* Ethnographic du Caucase is supplied by M. Michel 
Smirnov to the JRev. cFAnthrop. for April, 1878. 

K. 27 


former speak languages which are descended more or less directly 
from the Sanskrit V It would thus seem that the position of the 
Indian Dravidas is somewhat analogous to that of the European 
Magyars. Both have been assimilated to the Caucasic type, and 
both have accepted Aryan culture, while preserving intact their 
non-Aryan speech. 

The hirsuteness to which Peschel here refers occurs in a still 
more pronounced form amongst a few sporadic 
gups, such as the Todas of the Nilghen Hills, 
and the Ainus of Japan, who may be taken as living 
witnesses to the widespread diffusion of the Caucasic race through- 
out Asia in remote prehistoric times. Although now 
of Dravidian speech, the Todas (properly Toruwa 
" Herdsmen "), are distinguished from all the surrounding popula- 
tions by their splendid physique, perfectly regular Caucasic 
features, black wavy hair, full flowing beard, aquiline nose, light 
brown complexion and tall stature (5 ft. 9 in.) 2 . With the fodas 
de Quatrefages groups both the Kubus of Sumatra and the Ainus 
of North Japan. But the Kubus, although called "hairy men" by 
Col. Versteeg 3 , must be removed from this connection, for those 
seen and figured by Mr H. O. Forbes 4 exhibit no such peculiarity, 
while on the osteological evidence Dr J. G. Garson declares them 
to be "decidedly Malays and therefore Mongoloid 5 ." 

Not so the " Hairy Ainu," as they are correctly called by Mr 
A. H. Savage Landor 6 , one of the many observers 
who have described these aborigines of North-east 

1 Races of Man, p. 451. This view is fully in accordance with the now 
fairly established assumption that Aryan culture spread gradually southwards 
by a process of infiltration, resulting in a general fusion of the northern Hindus 
with the pre-Aryan peoples of the Deccan and Ceylon. Here again the slight 
Aryan-speaking element plays the part of the leaven in raising the aborigines to 
a higher plane of culture. 

3 W. E. Marshall, A Phrenologist amongst the Todas, 1873, passim. 

8 "Ce savant les appelle homines a poil et dit qu'ils sont entierement velus" 
(Races Humaines, If. p. 468). 

4 On the Kubus of Sumatra, in Jour. Anthrop. Jnstit. April, 1884, 

p. 121. 

6 Ibid. p. 132. 

B Alone with the Hairy Ainu, 1893. 


Asia at first hand in recent years. Although now confined to 
Yezo, part of Sakhalin and the southern members of the Kurile 
Archipelago, their territory appears to have formerly comprised a 
great part, if not the whole of Japan 1 , besides large tracts on the 
opposite mainland. In the national traditions there was a time 
when they could look out^on their watery domain, and exclaim, 
" Gods of the sea, open your divine eyes. Wherever your eyes 
turn, there echoes the sound of the Ainu speech 2 /' a speech now 
current amongst scarcely 20,000 full-blood and half-caste survivors 


of this remote Asiatic branch of the Caucasic division. Despite 
the attempts of some writers to affiliate them to the surrounding 
Mongoloid peoples, their claim to membership with the Caucasic 
family is placed beyond doubt by a study of their physical charac- 
ters. The features are not only regular in the European sense, 
but often quite handsome, with large slightly curved nose, clear 

1 <f From the relics of the Stone Age and of the Kitchen middens of Japan, 
Professor Milne concludes that the Amos once inhabited Japan as far south as 
Kiushiu" (Romyn Hitchcock, The Ainos of Japan, Washington, 1892, 

P- 455)- 

2 Quoted by the Rev. John Batchelor, The Ainu, of Japan, 1892. 



brown or greenish eyes set straight in the head, and olive brown 
or fair complexion. But the most striking trait is the abundance 
of coarse black wavy or crisp hair on head, face and body. " The 
Ainos are characterized by a strong growth of hair about the legs 
and body, long black hair on the head, and heavy beards 1 ." The 
type, however, varies, and those of Yezo differ considerably from 
the Tsuishikari Ainu of Sakhalin, while the low stature (5 ft. 2 or 
3 in.) and the skull of all shapes, long, round and intermediate, 
seem to betray secular interminglings with the neighbouring 
Mongoloid peoples 2 . 

For the INDONESIANS, ramifying to the left of our Family Tree, 
see Chap. XL 

1 Hitchcock, op. cit. p. 440. 

2 " De cet ensemble de donnees, on doit, ce me semble, conclure que les 
Ainos sont une race fondamentalement blanche et dolichocephale, plus ou 
moins alter6e par d'autres elements ethniques dont un, au moins, est essentielle- 
ment mongolique " (De Quatrefages, op. cit. II. p. 467). 


(pp. 100 et 

Recently the discussion as to the relative age of chipped and 
polished implements has taken an unexpected turn. Hitherto it has 
been taken for granted that, in the New World as in the Old, palaeo- 
lithic necessarily antedates neolithic culture, and that, where there is a 
time sequence, the chipped stones, being of ruder and simpler forma- 
tion, naturally precede the more perfected polished objects. But 
these views, which seemed placed beyond discussion, are now ques- 
tioned among others by Mr J. D. McGuire, who argues (The Archao- 
legist* July, 1894 ; The American Naturalist, January, 1895) tnat tne 
art of polishing by friction is easier and therefore antecedent to the 
flaking process. Mr Charles H. Read, of the British Museum, had 
little difficulty in exposing Mr McGuire's fallacies in a paper On the 
Evolution of the Art of working in Stone, in the American Naturalist 
for December, 1894 ; and little more would probably have been heard 
of the subject, had Mr McGuire not found a sort of "Advocatus 
Diaboli" in the distinguished ethnologist, Mr W. H. Powell, of the 
Smithsonian Institution. Referring to the researches of Mr Holmes, 
in the old stone workshops of the North American aborigines, this 
observer writes (Stone Art in America, American Anthropologist, 
January, 1895) : " In view of these facts, abundantly demonstrated far 
and wide over the continent, many American archaeologists and 
geologists have reached the conclusion that the distinction between 
* palaeolithic man '- and ' neolithic man/ as determined by the method 
of making the implements, is not valid for this continent. If these 
facts or the conclusion flowing from them startle European observers 
in geology and archaeology, it behoves them to reexamine their own 


facts, and if by the new methods of geologic observation they can 
demonstrate a time distinction between exclusively chipped imple- 
ments and mixed implements fashioned by both processes, we shall 
not fail to accord belief to their conclusions; but we shall hold the 
question open until assured that the new methods have been tried." 

Thus the war is, so to say, carried into the enemy's camp, and 
European archaeologists are asked to reconsider their own conclusions 
on a point about which no serious doubt has ever been raised. 
Mr Powell contends that because, for instance, the Shoshone Indians 
prepare their implements by chipping, and their Pahvant neighbours 
by rubbing, while the Uintahs employ both processes, there is no 
distinction between the palaeolithic and neolithic cultures, no time 
sequence, in North America, and consequently the same may, on 
further inquiry, be found to be the case also in Europe and elsewhere. 
The reply is obvious. There is necessarily a time sequence wherever 
the two cultures have been developed, whether in Europe, America, or 
any other part of the world. But this sequence does not prevent 
overlapping, the survival of primitive amid later methods, as fully 
explained at p. 72. In all cases the ruder precedes the more improved 
art, and under certain conditions may go on simultaneously with it, or 
even to its entire exclusion, as in Australia, Fuegia, or Somaliland. 
Speaking of the relations in the last-mentioned region, Dr Jousseaume 
remarks that the rude character of the flints is no sure test of their age. 
Those who have seen how the natives of the arid lands skirting the 
Red Sea are satisfied with the strictly necessary without seeking for 
artistic refinements, how their wants are limited to the point of 
privation, how under pressure they grasp the first rude implement at 
hand, will understand why " they have made no progress in working 
their flints, and why the art has remained rude (grossttre) throughout 
the stone age of those regions" (LAnthropologie, July-August, 1895, 
p. 411). Here we have a people still turning out rude palceoliths, 
while their Abyssinian and Galla neighbours, as well as some of their 
own kindred, have long been trained to the use of firearms. But what 
the advocates of the new theory have to show is that the firearms may 
be as old as the palaeoliths. Until this is done, no European archaeologist 
will ever believe that the polished implements of Mr Powell's Pahvants 
are absolutely as old as the chipped stones of the Shoshone Indians. 
The time sequence between the two cultures is merely obscured by the 
overlappings and survivals, by the intermingling of tribes at different 
stages of civilisation, and by the complete lack of historic records 
amongst illiterate populations with short memories and traditions 
going back at most to a few generations. 


To Mr McGuire's statement that polishing is easier and therefore 
older than flaking, it may be answered that much will depend on the 
nature of the material, whether flint, obsidian, soapstone, quartzite, 
sandstone and so on. But in any case it is a fallacy to suppose that 
the easier process necessarily comes first. Transport by wheeled 
vehicles or by steam is immeasurably easier than by pack animals or 
by porters ; yet these coire first in the order of evolution, and all 
labour-saving methods are a distinct mark of progress. 

Attention may here be called to Mr Thomas Wilson's Primitive 
Industry, where it is shown that Mr Holmes's researches (see above) 
prove nothing against Dr Abbott's Trenton gravel finds (p. 101), which 
are regarded as on the same time level as those of the European 
palaeolithic deposits (Annual Report of the Smithsonian Institution 
for 1892, Washington, 1893, p. 321). 

Still more important in connection with this subject is the paper 
on the Antiquity of Man in North America^ contributed to the Ame- 
rican Naturalist for June, 1895, by Mr E. D. Cope. This eminent 
palaeontologist frankly admits palaeolithic man in the northern con- 
tinent ; and, like Mr Wilson, denies that the question is affected by the 
investigations of Mr Holmes. He has himself collected some obsidian 
spear-heads in a deposit in Oregon in association with an extinct 
fauna, which he holds to be contemporary with the pleistocene fauna 
represented by Megalonyx^ Mylodon and other fossil remains on 
the east side of the continent. In the face of evidence such as this 
the persistence of primitive cultures side by side with later develop- 
ments loses all significance. If these cultures are now synchronous in 
some regions, clearly they were not so ab initio; palaeolithic must have 
preceded neolithic processes in America as elsewhere, unless we are 
prepared to admit the possible existence of neoliths as well as palaeo- 
liths in association with an extinct pleistocene fauna in the northern 
continent. But in that case cadit quasUo, and chipped and polished 
implements will all alike have to be regarded as dating back not 
merely to prehistoric times (McGuire, Powell), but to the pleistocene 
epoch, as in Europe and other parts of the Eastern Hemisphere (Cope, 

(P- 93)- 

The range of primitive man is now extended to the Irawadi basin, 
Indo-China, by the discovery of chipped implements in some tertiary 
deposits near Yenangyoung on the Irawadi, Upper Burma (29 21* 


N. lat.) by Dr Fritz Noetling, of the Geological Survey of India, in the 
year 1894. At first the beds were supposed to be Upper Miocene, as 
stated by Prof. T. Rupert Jones in a reference to the subject in the 
Geological Magazine for November, 1894. But in reply to an inquiry 
by Mr W. T. Blandford, Dr Noetling afterwards explained that he 
had " definitely ascertained that the bed containing the chipped flints 
is Pliocene " (Nature, April 25, 1895, P- ( 8 )- A ful1 account of the 
discovery was given by the finder at a meeting (Oct. 20, 1894) of the 
Berlin Gesellschaft fur Anthropologie, Ethnologic, und Urgeschichte. 
To Mr R. D. OldhanVs suggestion that the flints may have been 
found, not in situ, but in the coating of mud washed down from the 
superincumbent strata (Natural Science, Sept. 1895), Dr Noetling 
replies convincingly in the same periodical, April, 1897. 

(pp. 149150). 

M. Ph. Salmon gives final results as under : total number of skulls 
measured, 688; of which 577 per cent, are classed as dolichocephalic ; 
24*1 mesaticephalic with index ranging from 77 to 79; 21 '2 brachy- 
cephalic ; most frequently recurring index 73 (Revue mensuclle de 
VfLcole d' Anthropologie de Paris, May 15, 1895). 

(pp. 179, 182; and CHAP. XI.). 

Two skulls from Croydon in North Queensland recently added to 
the collection in the University Museum, Cambridge, are described by 
Mr W. Lawrence Henry Duckworth in the Journal of the Anthro- 
pological Institute for February, 1895, pp. 213218. The first, that 
of a male adult, is marked by massive, overhanging brows, large 
upper jaw, strong malar bones and zygomatic arches, low cranial 
capacity (1,255 c c -)> extreme dolichocephaly (687) and pronounced 
prognathism in norma lateralis, though this feature is not brought 
out by the gnathic index (96*9). The second, an adult female, shows 
"much general similarity" to the other, with capacity 1,205 > cephalic 
index 69-9; gnathic index 77-3. "To select the characteristics of 
the pair would be to emphasize : (i) the prognathism, (2) the great 
vertical height from basion to bregma, (3) the shallowness of the 


glenoid fossa. Of these the marked prognathism is interesting from 
the fact of the same characteristic distinguishing Melanesian skulls ; 
the same may be said of the basi-bregmatic height. As regards this 
latter, the result is a height index greater than a breadth index. 
Such a condition is common in Melanesians, common in skulls from 
the more northern parts of Australia, but progressively rarer as one 
advances to the south" (p. 115). 



Six skulls from the East coast of Labrador lately presented to 
the Cambridge University collection by Dr E. Curwen, and described 
by Mr W. L. H. Duckworth in the Journal of the Anthropological 
Institute for August 1895, show an extreme degree of dolichocephaly 
with cephalic index ranging from 75*4 to 65-8. On the other hard 
the cranial capacity is high, rising from 1340 and 1385 to 1480 and 
1550, and in one instance to 1790 c.c. In general the principal 
measurements and indices "depart in no very important points from 
those already recorded by other observers " (p. 72). 

(p. 170). 

Fresh evidence on this assumed character of the hair of the dark 
races is supplied by Prof. Virchow's paper on the Dinkas of the 
White Nile in Zeitschrift fur Ethnologie, 1895, Heft II. p. 152. "The 
hair possesses uniformly that property which I have so often described 
under the name of ' spiralgerollt ' [the corkscrew twist]. The so- 
called 'peppercorns' arising from the closeness of the twist develop 
in the longer growths those thick curly locks to which is mainly 
due the * woolly ' look of the hair. Between these peppercorns there 
occur apparently bald spaces which again gives the impression 
that all the hair forming each grain grows from a single spot. If the 
hair be left uncut, so as to acquire a certain length, as is often the 
case with the women, the grains dispose themselves in continuous 
rows, giving rise to long ridges with intervening empty spaces, as 
in the artistic arrangement on a hairdresser's dummy. But such 
spaces are no more hairless than are the partings made by combs, 
so that the hair forms no separate clusters or tufted growths 
('Buscher). The process appears very early in life, as in the ten 
months' old child of Am61, a member of the Req tribe." 



(CHAP. IX.). 

In a paper on Pithecanthropus Erectus, contributed by Dr Arthur 
Keith to Science Progress for July 1895, Prof. Cunningham's conclusion 
that the Java remains indicate "a human race more primitive than 
any hitherto discovered," is accepted as "very probably right." From 
a study of the facial parts specially modified for speech Dr Keith 
further infers that, the arrangement of the mental lines being the same 
in human fossil jaws as in modern ones, "the muscles which arose 
from them were adapted to similar purposes, and were therefore 
subservient for speech. The arrangement of the mental lines in 
anthropoids is quite different. They turn up in front of the inferior 
canine teeth, and enclose between them a quadrilateral rough surface 
corresponding to the triangular mental space of man. In anthropoids 
this space retreats rapidly downwards and backwards, a feature in 
which fossil man resembles apes much more than modern man, and 
shows also, I think, that fossil man was less highly adapted for 
speech" (p. 364 5). From this it appears that primitive man was not 
speechless, but that his articulation was less perfect than that of his 
modern descendants. Thus is confirmed on anatomical grounds our 
statement that "speech is a function which perfects itself hand in 
hand with the growth of the organ. Hence the faculty starts from 
a germ, and its history is one of continuous upward evolution from 
slowly accumulating crude utterances" (p. 195). It follows as a 
necessary corollary that the organic or present condition of speech 
was preceded by an "inorganic phase" as shown in our diagram, 
p. 212. 

(p. 179). 

Supply Topinard's formula :- ^.. That is to say 

the cephalic index of breadth, as distinguished from the less important 
cephalic index of height, is found by multiplying the maximum trans- 
verse diameter by 100 and dividing by the maximum antero-posterior 
diameter, as described at p. 178 9. 


Abbeville, flints found at, 82 
Abbott, C. C., on the flint-bearing 

Trenton gravels, 101 
Abydos district, Nile Valley, flints 

found in, 92 
Abyssinia, a region of ethnical con- 

fusion, 390 
Abyssinians, their constituent ethnical 

elements, 390 i 
Adelaide tribe, South Australia, of 

Neanderthal type, 238, 294 
AetaNegritoes,Philippine Is. ,25961; 

their physical characters, 260 ; men- 

tal capacity, 261 
Afars, their ethnical relations, 386 

_ 7 

Afghans, of Irano-Semitic type, 397 
African Negritoes, 246 53 
Agglutination, many kinds of, 205; 

its nature, 209; the first phase of 

organic speech, 209; passes into 

Inflexion, Polysynthesis and Isola- 

tion, 211 2 

Agricultural state, see Social States 
Ainus, their physical characters and 

ethnical relations, 418 20 
Akkads of Babylonia, type and speech, 

Ak-kapana, its megalithic doorway, 


Akkas, early records of, 245 ; their 

type, 247 
Alfuro, a term of no ethnical value, 

note, 328 
Alignments, monolithic monuments, 

ophylian whites," 308, 318 
ghino, S., on tertiary man i 

in S.* 


America, 98 
America, its prehistoric monuments, 

106, 138 9; accessible from Europe 

in Miocene and later times, 231; 

and from Asia at all times, 232; 

peopled during the Stone Ages, 341 
2; later migrations disproved, 


American aborigines, their relation to 
the Mongol group, 222, 336; to 
the black and white groups, 337 
40; aberrant types explained, 337 
9; uniform types of, 338; not de- 
rived from historical Asiatic peoples, 
342> 365 7; two types, long and 
round-headed, 346 7; everywhere 
intermingled, 347 9; their physi- 
cal and mental characters compared 
with the Mongolic and Ethiopic, 
349 52 ; their temperament every- 
where uniform, 353 7; reached 
America by two routes during the 
Stone Ages, 362; present distri- 
bution of the two types, 362 ; theii 
various destinies in Latin and Anglo- 
Saxon America, 370 2 

American culture of independent 
growth, 340; evidences of foreign 
influences disproved, 340 2 ; 
identical with that of the Old World 
in the first Stone Age, later diver- 
gent, 3456 

A merican languages, their morphologv, 
2ii 3; their uniform character, 
357 **; absolutely distinct from 
those of the Old World, 358; great 
number of American stock langua- 
ges, 359 6 

Amiens, flints found at, 82 

Amzigh, national name of the Mauri- 
tanian Berbers, note, 384 

Anchitherium, 37 

Andamanese Negritoes, their physical 
and mental characters, 256 7 

Angkali, "Fishing Chukchi," 319 

Anglo-Americans, their physique due 
to convergence, 372 3 

Annamese of Indo-China, 321 



Anthropoidea, the five families of the, 

i7 *8; generalised anthropoid form 

(diagram), 19 
Anthropology* special and general, 

defined, i; scope of, i, 2; criminal, 


Anthropomorphism, its origin, 217 
Anthropopithecus (Chimpanzee), 18; 

A. sivalensis, 18 
Antiquity of man, 50; not a question 

of orthodoxy; 75; see also Palaeo- 
lithic and Neolithic 
Arabs, their ethnical relations, 393; 

their type, 313 4J language, 304 
Argentina, evidences of primitive man 

in, 98, 99 
Armenians, of Semitic type, 397; 

their language, 411 
Aryan, meaning of the term, 227 ; note, 

395; a linguistic rather than an 

ethnical expression, 3956 
Aryan languages, their wide diffusion, 

410; their main branches, 411 2; 

their disintegration, causes of, 412 


Aryan problem, summed up, 410 
Aryan root theory exploded, 208 
Aryans, their migrations, 396-~9; 
their cradleland, 400 5; primi- 
tive culture, 401 2; early contact 
with the Semites, 404 5; their 
physical characters, 405 10 
Ashton, W. G., on Koreo-Japanese 

linguistic affinities, 313 
Asia Minor, palaeolith found in, 93 
Assyrians, their ethnical relations, 

393 * language, 394 
Austral secondary and tertiary con- 
tinent, 230 ; probable cradleland of 
man, 236 
Australia, evidences of quaternary 

man found in, 94 

Australian languages not related to 
Dravidian, 233; nor to Malayo- 
Polynesian, note, 292 
Australians, their cerebral structure, 
47 ; not a homogeneous group, 289 ; 
their physical characters, 290 i; 
their constituent elements, 291; 
their relations to the Tasmanians 
and to Pliocene man, 292 4 ; their 
craniology, 424 

Auvergnats, their ethnical relations, 406 
Avars, a Finno-Tatar people, 308 
Avery, J., on the Tibetan language, 


Aymaras, their culture, 139 40 

Bacon, Friar, on the experimental 

method, 41 
Bacon, Nicholas, his inductive method, 

Bahrein Islands, their sepulchral 

mounds, 134 
Ball, V., on the Negrito question in 

India, 256 
Baltic ^inns, 311 
Bantu, origin and meaning of the 

term, 271 2 
Bantu negroid peoples, their type 

and range, 271; their speech, 272 

3; chief tribes, 277 80 
Bantu prefix particles, note, 271 2 
Barrows, their construction, 124; 

origin of the word, note, 124 
Bartram, H., on the independent 

character of American culture, 343 
Basque language, its morphology, 213; 

its Berber relations, 205, 377 
Basques, their relations to the Picts and 

Berbers, 377 9 5 their type, 378 
Batwa Negritoes, 248 
Beard, a racial character, 177 
Beddoe, Dr, on reversion, 35 
Bejas, their ethnical relations, 387 
Bent, Theodore, on the Bahrein 

monuments, 134 
Berbers, their relations to Neolithic 

man, 376; to the Basques, 377 8; 

of black type, 382 ; their relations 

to the Blemmyes, 386 
Bernier, F.,his primary human groups, 


Bimana, the, 17 
Binger, Capt., on the early closing of 

the cranial sutures in the Negro 

race, 44 
Blemmyes, their ethnical affinities, 

Blumenbach, on the specific unity of 

man, 160, 161; his primary human 

groups, 164 
Blumehtritt, Prof., on the affinities of 

the Philippine Islanders, 332 
Blyden, Dr, a Negro writer, note, 

Boas, Dr Fr., his theory of the tribe, 

7; on the Indian half-breeds, 152 
Bonneval dolmen, note, 127 
Bory de Saint- Vincent, his primary 

human groups, 164 
Boucher de Perthes, his researches in 

the Somme Valley, 75, 82 
Bourgeois, Abbe, champion of Ter- 
tiary man, 75 



Bove, G., on the Chukchi Hyper- 
boreans, 319 

Boyd-Dawkins, Prof., on Pre- and 
Post-glacial man, 63; on Lyell's 
Seasonal-migration Theory, 64 ; on 
the continuity of Palseolithic and 
Neolithic cultures, 73, 113 

Brachycephaly, defined, 178 

Bradley, H. , on the Gothic language, 
note, 399400 

Brain, weight of, in man and the lower 

orders, 40, 41; its cellular tissue 
(grey cortex) scat of mental energy, 
44; its comparative study yields 
better results than craniometry, 47 

Branch, meaning of the term, u 

Brassempouy Palaeolithic station, its 
ivory statuettes, 252 

Brazil, evidences of primitive man in, 
98, 99 

Breal, M., on pre- and post-positive 
languages, note, 214 

Brinton, Dr D., on the assumed Asi- 
atic origin of American cultures, 
2iJ> 9; on evolution per saltum, 
235; on the cradleland of the 
Semites, 392 

Brittany, its Neolithic monuments, 
129, 133 

Brixham Cave, implements found in, 

Broca, Paul, his definition of Anthro- 
pology, i ; his theory of the French 
Kelts, 34; his cranial measure- 
ments, 43, 46; turns from cranial 
to cerebral studies, 47 ; his primary 
human groups, 168; his scheme of 
cephalic index, 179; on the brain of 
man and the apes, 191, 192 

Bronze, origin and diffusion of, 123; 
not unknown in America, 123 

Bronze Age, 56, 109; in Peru (Chimu), 

123. 335 
Brown, J. A., on the continuity of 

primitive cultures, note, 115 
Briix, human remains found at, 146 
Bubi, their type, note, 278 
Buckland, Miss A., on the origin of 

the Nuraghi, 126 

Buddhist purgatory and Aztec spirit- 
land, 218 
Buffon, on the natural history of man, 

Bulgarians, their ethnical and linguistic 

relations, 309 10 
Burma, flints found in pliocene beds, 


Burmans of Indo-China, 321 

Burmeister, on early man in Argentina, 

Bushmen, their relations to the Hot- 
tentots and Bantus, 249; their 
mental qualities, 249 50; present 
social state, 250 

Byrne, Dean, on the correlation of 
speech and temperament, note, 357 

Cae Gwyn Cave, flints found in, 81 
Cairns, their construction, 124; origin 

of the word, note, 124 
Cairo, implement founr 1 near, 92 
Calaveras, fossil skull found at, 104 
Calendars, American and Asiatic, 218 
Callard, T. K., on the antiquity of 

man, 76 
Canada, implements found in, 101, 

Cannibalism in the Congo Basin, 255; 

in Hayti, 267 
Cannstadt skull, its evidence doubtful, 

Canterbury gravel beds, eoliths found 

in, 82 

Cape Colony, palaeoliths found in, 93 
Carles, his discoveries in Argentina, 

34 . 

Carnac, its menhirs, 133 

Carus, Dr, on the correlation of phy- 
sical and mental characters, 47; his 
primary human groups, 165 

Castenedolo man, 32 

Catlin, G., on the endurance of the 
M and an Indians, 354 

"Cattle-Herd Theory," the, 14 

Caucasians of dark type, 381 

Caucasic, meaning of the term, 226; 
its use justified, 226 

Caucasic element in Polynesia, 289; 
in the Mongol domain, 297 9, 

Caucasus, human fossils found in, 93 ; 
inhabitants of, their ethnical and 
linguistic relations, 415 7; chief 
divisions, 415 6 

Cebidae, the, 18 

Cephalic index, how determined, 178, 
179, 426; tables of, 179, 180 

Cercopithecidse, the, 18 

Chalk Plateau, Kent, eoliths found on, 

Chapman, F. R., on the Stone Ages 
in New Zealand, 95 6 

Charancey, M., on the Otomi lan- 
guage, 214 



Chelles, flints found at, 84 

Chellian Age, 84 

Chibcha language, its monosyllabism, 

Chichen-Itza, 138 

Chillicothe, its monuments, 107 

Chimpanzee, species, range and charac- 
ters of, 22, 24 

Chimu, Peru, bronze art of, 123, 335; 
its pre-Inca culture, 335 

Chinese language, originally polysyl- 
labic, 207 

Chinese race, its physical and mental 
characters, 321 2; estimates of 
population, 322 

( hudes, "white-eyed Finns,'* 307 

Chukchi, their ethnical relations, 318 

Chuklukmiut, Eskimo of N.E. Sibe- 
ria, 318 

Ciscvaens, their origin and construc- 
tion, 124 

Clan, defined, 5, 6; clan system of 
kinship, 6 ; derivation of the word, 
note, 8 

Claymont, Delaware, implements 
found at, 101 

Ciemence Royer, Mme., on glacial 
phenomena, 63 

Clough, J. C., on mixed languages, 

Codrington, Rev. R. H.,on the Mela- 
nesian languages, 287, 331 

Colour of the skin, a racial test, 171 ; 
of Negro children, 1 73 

Colours, racial, the six primary, 173 

Comhaire, C. J., on a bronze age in 
Belgium, 109 

Cope, E< D., on the man of Spy, 35, 
147 ; on the evolution of the An- 
thropomorpha, 35; on palaeolithic 
man in North America, 423 

Coptos, age of its rude monuments, 

Corancez dolmen, note, 127 

Couvade, widely prevalent, 219; its 
explanation, 368 

Cranial capacity in man and the 
Simiidse, 40; to be distinguished 
from mental capacity, 42 ; of various 
races, 43; correlated to mental 
capacity, 190; its relation to the 
weight of the body in the animal 
series, 191 ; in man and the higher 
apes, 191 ; comparative tables of, 
192; American inferior to Mongol 
and Negro, 352 

Cranial indices, 43, 46; Negro, 264; 
Malay, 328; American, 349 

Crannogs, Irish and Scotch, 122 

Creswell Caves, implements found 
in, 80 

Criteria of race, physical, 171 ; mental, 

Croll, Dr, his periodicity theory, 57 

Cro-Mr^non race, 149; its ethnical 
relations, 376 

Cromlechs, their origin and con- 
struction, 124 

Cruel, Dr R., his "Turanian" theory, 
note, 406 

Cultures, grades of, often synchronous, 
hence not always a test of time se- 
quence, 72; Palseo- and Neolithic, 
comparative tables of, no, in; 
continuity of, in, 112 

Cunningham, Dr J., on Pithecanthro- 
pus erectus, 145 

Customs, as racial tests, see Usages 

Cuvier, his multi-creation theory, 30 ; 
his primary human groups, 164 

Cycloliths, 130 

Cyrus, Thomas, on the mound-build- 
ers, note, 107 

Dallas, J., on the dispersion and mi- 
grations of primitive man, 233 4 

Damaras, 253 

Danakils, their ethnical relations, 

Dano- Eskimo half-breeds, permanent- 
ly fertile, 152 

Darwin, his view of race and species, 
5 ; his explanation of rudimentary 
organs, 29 ; on species and variety, 

Dauertypus, Kollmann's, 34, 36, 37 

Davis, Dr Barnard, on reversion, 35, 
his table of cranial capacity, 192 

Decle, L., on the Waruanda, 389 

De Lapouge, Prof., on Aryan ethnical 
relations, 407 

De Mortillet, his four Palaeolithic 
epochs, 83, 84 

De Nadaillac, on the peopling of 
America, 342 ; on the Aryans, note, 

39. 6 

Deniker, J., on the West African Ne- 
groes, 153; on pure and mixed 
races, note, 163; his scheme of 
classification,. 168, 169 

Denmark, its peat-beds, 118; its kit- 
chen middens, 119; its neolithic 
monuments, 135 



Dentition, human and simian, 25, 26, 
1 84 ; its correlation to gnathism, 1 84 ; 
cause of defective in civilised man, 
184 ; its relation to social conditions, 

De Quatrefages, his theory of evolu- 
tion, 31 ; on the Thenay and Monte - 
Aperto finds, 91 ; on early man in 
S. America, 99 ; his classification of 
man, 167; his theory of early mi- 
grations, 232; on the evolution of 
the Hominidse, 239; on "black" 
and " white" American aborigines, 
337; on the relations of Neolithic 
man in Europe and Africa, 376; on 
the cradleland and dispersion of 
primitive man, 376 7 

De Ujfalvy, Ch., on the Galcha race 
and languages, note, 411 

Diagram of linguistic evolution, 212 

Dolichocephaly, defined, 178 

Dolmens, their origin and construction, 

Donnenberg, E. G., on the dwarfs of 
MaFocco, 247 

Dorsey, J. O., on the fate of the N. 
American aborigines, 37 1 2 

Douglas, R. K., on Akkadian and 
Chinese monosyllabism, 207 

Dravidian languages not related to 
Australian, 233 

Dravidians, their relations to the Aus- 
tralians, 290^ their ethnical affinities, 
417 8; their chief divisions, 417 

" Druids' Altars," origin of, 125 

Dryopithecus, 18; relations of to the 
Simiidre and to man, 24 

Dubois, Dr E., his Pithecanthropus 
erectus, 144 

Dunn, Dr R., on mixed races, 155 

Dupont, E. , on the pleistocene fauna, 


Duval, M,, on organ and function, 42 
Dwarfs in Marocco, 247; formerly 

widespread, 246 

Earl, W., on grades of culture in 
Malaysia, 216 

Eastern Polynesians, a branch of the 
Indonesians, 326 

Egede, Pastor, on the temperament 
of the Eskimo, 353 

Eguisheim, fossil skull found at, 148 

Egyptian culture, antiquity of, 56; 
ethnical and linguistic relations, 391 

Elizabeth Island, Fuegia, its shell- 
mounds, 96 

Ellis, A. B., his explanation of the 
Clan system, 6 ; on the early clos- 
ing of the cranial sutures, 44 

Emin Pasha, his measurements of the 
Akka dwarfs, 189 

Endogamy, meaning of the word, 
note 2, p. 7 

English, not a mixed language, 199; 
its morphology, 213 

Eohippus, 36 

Eoliths, British, 74; found at Finch- 
ampstead, Paddington &c., 82; 
derivation and meaning of the 
word, note, 82 > 

Equidse, their evolution compared 
with that of the Hominidae, 36, 37 

Erdeven, its alignment, 131 

Erith, river-drift implements found 
at, 82 

Eskimo of Siberia, 318 9; of Ame- 
rica, their temperament, 353; thoir 
ethnical relations unravelled, 363 
5 ; their type, 363 4 ; their cranio- 
logy, 4^5 

Esthonians, Baltic Finns mentioned 
by King Alfred, 311 

Ethiopic, twofold meaning of the 
term, note, 227 

Ethnography, defined, 2 ; scope of 

2 > 3 
Ethnological nomenclature, 3, 4; 

definite terms, 4 12; indefinite 

terms, 13, 14; example, 15 
Ethnology, defined, 2 ; scope of, 3 
Eugenesis, defined by Broca, 153; 

proved for all races, 155 
Eurafrican miocene continent, 230 
Eurasian steppe, cradleland of the 

Aryans, 403 
Eurygnathism, 181 
Evans, A. J., on the culture of the 

Mentone Cave Men, note, in 
Evans, Sir J., on the gap between the 

Old and New Stone Ages, 112 
Evolution, physical and mental of 

man, 16 49; "persaltum," 235 
Exogamy, meaning of the word, note 

2, p. 7 
Eye, colour and shape of, racial 

characters, 186; more uniform in 

the lower than in the higher races, 

186; the Mongol and Semitic, 187 

Face, its general expression a racial 
character, 187 ; four types of, 187 

Family, twofold meaning of the word, 
8; the social unit, b; origin of, 



conflicting theories, 8, 9; family 
life amongst the Anthropoidea, 9 

Family Tree of the Hominidse, 224 ; 
of Homo ^Ethiopicus, 244 ; of 
Homo Mongolicus, 300; of Homo 
Americanus, 361 ; of Homo Cau- 
casicus, 380 

Fans, their type and speech, note, 278 

Finchampstead, eoliths found at, 82 

Finger markings, rather an individual 
than a racial test, 189 

Finn, meaning of the word, 305 

Finno-Tatars, their early and later 
migrations, 308 u 

Finno-Turki, divergent types, 305 

Finns, their Mongolo-Caucasic affini- 
ties, 305 8; Baltic, 307; their 
cradleland, 308 

Finsch, Dr O., on the Maori, note, 
327 ; on the Malays, 330 

F:re, reminiscences of its origin, note, 

Flores, former presence of Negritoes 
in, note, 261 2 

Flower, SirW., on reversion, 35; on 
the Negrito and Papuan crania, 1 78 

Flower and Lydekker, their primary 
human groups, 169, 221; on the 
relation of the American aborigines 
to the Mongol group, 221 2 ; on the 
dispersion of the Hominidae over 
the globe, 240; on the Australian 
problem, 291 2; on the physical 
characters of Homo Caucasicus, 382 

Food, of early man and the higher 
apes, no, in 

Forbes, H. O., on the Sumatran 
Kubus, 418 

Fort Ancient, Ohio, its earthworks, 

Franco-Canadian half-breeds, perma- 
nently fertile, 152 

Fuegians, their mental capacity, 48 

Fulahs, physical type, 270; their 
political ascendancy, 270 

Function and Organ correlated, 41, 

Gabelenz, G. von der, on the Basque 
and Berber languages, 205, 377 

Gafsa, Tunisia, implements found at, 

Gaillard, F., his theory of the mega- 
lithic monuments, 137 

Galcha languages, 4101 

Galchas, their type and ethnical rela- 
tions, 405 6 

Gallas, their type and ethnical rela- 

tions, 387 8 
Galley Hill gravels, human remains 

found in, 147 

Gal ton, Fr., on reversion, 35 
Garcilaso de la Vega, on the Tiahua- 

naco monuments, 139 
Garner, R. L. , on ape language, note, 

Garson, Dr J. G., on the Sumatran 

Kubus, 418 
Gatschet, A. S., on the American 

polysynthetic languages, 214 
Geoffrey Saint-Hilaire, his four fun- 

damental types, 1 66 
Geological sequence, the, its bearing 

on man's antiquity, 50; table of, 

Geological time, various estimates of, 

5' 58 
Germans of Caucasia, their changed 

type and unchanged speech, 203 
Germans, primitive and later types 

of, 397 

Gesture language, 195 
Gibbon, 18, 20; range and species of, 

Giglioli, E., on inter-glacial man, 70; 

on primitive Australian implements, 

note, 95; on the Australian pro- 

blem, 293 ; on the Afars, 387 
Glacial problem, the, 56 ; restated, 68 
Gnathism, a racial character, 181; 

various grades of, 181 ; their evolu- 

tion, 181; facial, 181 ; sub-nasal, 

181, 182; table of, 182 
Gomme, G. L., his theory of the 

human horde, 1 3, 14 
Gonaquas, 253 

Gorilla, range and characters of, 22 
Gothic language, its late survival, 

399 400 
Gray, J. , on the relations of Picts and 

Iberians, 378 9 
Greeks, their primitive and later 

types, 4089 
Green, T. H., his theory of the 

family, 9 
Griquas, 253 

Grotte de la Vache, its culture eras, 90 
Guatusos of Costa Rica, legendary 

statements regarding their origin, 


Guillemard, Dr F. H. H., on the 
Negritoes of ihe Philippine Is., 261 ; 
on the colour of the Japanese, 316 ; 
on the morais of Polynesia, 370 



Gyarungs of Central Asia, their 
Oceanic affinities, 326 

Haeckel, his Homo primigenius 
alalus, 30 ; his classification of man, 

Hahn, Dr E., his six Kulturformen, 

Hair, a test of race, 1 74 ; its 3olour 
and texture, 174, 175; threefold 
structure of, 1 76 ; tufted non-exis- 
tent, 425 

Hairy men, reports of, 26 

Hale, Horatio, his theory of the 
family, 8 ; on the relations of 
speech to anthropology, 193 

Hallstadt necropolis, 109 

Hamites, white, brown> and dark, 
383 ; Eastern division of, 586 9 ; 
of 'Abyssinia, 390; relations to the 
Semites, 391 

Hamitic languages, three groups, 391 

Hamito-Semitic, ethnical and linguis- 
tic Affinities, 391 

Hamy, Dr, on the physical characters 
of the Sudanese Negroes, 269 ; on 
Mongolo-Caucasic interminglings, 
note, 299 ; on the Indonesians, 326 

Hapalidse, the, 18 

Harratins, " black Berbers," 382, 

Harris, W. B., on the Berbers of 

Marocco, 383 

Hatfield Beds, flints found in, 78 
Hayti, Negroes of, 267 
Height, see Stature 
Hellenes, their primitive and later 

types, 408-9 
Herbert Spencer, on the brain of 

civilised man and the savage, 191 
Himyarites of Yemen and Abyssinia, 

Hindus, their ethnical relation, 398 

9 ; their primitive type, 409 
Hipparion, 37 

Historic period, its duration, 56, f 16 
Hitchcock, Romyn, on the Ainus, 

note, 419 
Hiung-nu, a historic Turki people, 

Hodgson, B. H., on the relations of 

the Asiatic and Oceanic languages, 

Hohnel, Lieut, von, on the Masai, 

3 8 9 . 

Hominidse, the, 1 7 ; four primary 
varieties of the, 25 ; already diffe- 


rentiated in quaternary times, 
33, 34; Tertiary, rejected, 37; 
are varieties not distinct species, 
1 4 2 3 ; the physiological argument, 
142 ; the anatomical argument, 
156; the psychic argument, 160; 

. comparative table of their physical 
and mental characters, 228; their 
probable cradleland, 236; their order 
of evolution, 239; their dispersion 
over the globe, 240 

Hommel, Dr F., on the relations of 
the Hamites and Semites, 39 1 

Homo y^Ethiopicus, ideal type of, 224; 
two divisions, African and Oceanic, 
242; his Family Tree, 244; his 
early migrations, 245; see also 

Homo Americanus, ideal type of, 
224 ; relations to the Mongol group, 
336; uniform physical type, 336- 1 - 
7 ; Family Tree of, 361 ; see also 
American aborigines 

Homo Caucasicus, ideal type of, 224; 
dominant position of, 226; cradle- 
land of, 374 5 ; first migrations, 
375 6 ; Family Tree of, 380 ; 
physical characters, 382 

Homo Mongolicus, ideal type of, 224; 
aberrant types of, 226; original 
home of, 295; early migrations of, 
296; evolution of, 296; physical 
characters of, 297; see also Mon- 

Horde, derivation and meaning of the 
term, 13; implies no kinship, 13; 
its historic and ethnical use, 13, 14 ; 
theory of the human horde, 14 

Hor-soks, Mongolo-Turks of Tibet, 

Hottentots, their relations to the 
Bushmen and Bantus, 249; their 
mental and physical characters, 251; 
formerly widespread, 252 3; pre- 
sent range and position, 253; speech, 


"Hottentot Venus" of Cuvier, 251 
Houze, Dr, on heredity, 35 
Hovas, their modified Malay type, 

Hovelacque, A., his polygenist views, 

Howarth, O. H., on fancied traces of 

Japanese in America, 340 I 
Hoxne, palseoliihs found at, 78 
Hudson, W. H., on primitive man in 

Patagonia, 97 




Humboldt, A. von, on the tempera- 
ment of the Carlos, 355 

Humboldt, W. von, on the Basque 
language, 213 

Huxley, his primary human groups, 
167; his Xanthochroid and Mela- 
nochroid divisions, 379 80 

Hyperboreans of Siberia, 317 

Hypsistenocephaly, 180 

Ibero-Berber relations, 378 9 

Ilford, river-drift implements found 
at, 82 

Implements, their value as tests of 
antiquity, 76 

im Thurn, E., on the temperament of 
the Guiana Indians, 356 

Incas, the, destroyers not founders of 
Tiahuanaco, 139 

India, palseoliths found in, 93, 94 

Indian half-breeds, U. States, perma- 
nently fertile, 152 

Indo-African Miocene Continent, 229; 
probable cradleland of man, 236 

Indo-Europeani) A 

Indo-Germans ) * 

Indonesians, a sub-division of the 
Caucasic family, 326 ; their Ocean- 
ic domain, 326 ; their physical 
characters, 327 8 

Indo-Oceanic linguistic affinities, 326 

Inflection, nature of, 212; reverts to 
agglutination, 210; grows out of 
agglutination, 215; merges in poly- 
synthesis, 214 

Inter-glacial man, 63, 67 

Iranians, their primitive type, 407 

Jackson county, Indiana, implements 

found in, 101 
Jagor and Koerbin, on the low-caste 

tribes of S. India, 255 
James, F. L., on the Somali, 388 
Japan, artificial caves and other 

evidence of primitive man found in, 

Japanese, their relations to the 

Koreans, 313; their origin and 

type, 3145; their mental quali- 
ties, 316 7 
Jastrow, Prof., on the cradleland of 

the Semites, 392 
Java, remains of pleistocene man in, 

144; Negritoes of, 262 
Jennings, A. V., on the age of the 

Mentone Cave men, 73 

Jespersen, Prof., on the evolution of 

speech, 209 
Johnston, H. H., on the Bushmen, 

251; on the Negro temperament, 

Junker, Dr, on the hair of the 

Negroes, 175; on the Negritoes, 

248; on Negro and Bantu inter- 

mhiglings, 274 

Kabyles, their relations to Neolithic 

man, 376 

Kalang Negritoes, Java, 262 3 
Karelian Finns, 307 
Karons of New Guinea, 261 
Keith, Dr A., on the evolution of the 

organs of speech, 426 
Kelt, a term of no ethnical value, 


Kelts, the, not dolmen-builders, 125, 
136; their migrations, 136; their 
relations to Picts and Iberians, 378 
9; their multifarious types, 397, 

Kent's Cavern, implements found in, 

Khasi Hills, neolithic monuments in, 


Khasi language, 325 
Kiranti language of Nepal, 325 
Kirghiz, a Turki people of West 

Siberia, 312 
Kirkdale Cavern, 75 
Kitchen middens, no sure test of age, 

77; of Denmark, their age, 119; 

of Fuegia, their age, 96 
Kizil-bash Turks of Anatolia, 312 
Knight, E. F., on the temperament of 

the Paraguay Indians, 355 
Kolarian languages, 325 
Kolea, Algeria, flints found at, 93 
Kollmann, his Dauertypus, 34, 37; 

his anatomical argument for the 

unity of man, 156; on Neolithic 

Negritoes, 246 
Koranas, 253 
Koreans, their ethnical and linguistic 

relations, 313 

Koryaks of N.E. Siberia, 319 
Kubus of Sumatra, their ethnical 

relations, 418 
Kurgans, prehistoric structures, 126 

Lacouperie, T. de, on Chinese mono- 
syllabism, 207 ; on Chinese origins, 
323; on Asiatic and Oceanic liu- 
guistic affinities, 326 



La Denise, human remains found at, 

Lagoa Santa, human remains found 
in, 99 

Lahotan, Lake, palaeolith (?) found 
in, 103 

Lake Dwellings, of Switzerland, 120, 
122 ; of Ireland and Scotland, 122 ; 
wide range and origin of, 12 * 

Laloy, L., on the West African Ne- 
groes, 153 

La Madeleine Rock Shelter, imple- 
ments found in, 87 

La Naulctte, fossil skull found at, 145 

Landor, A. H. Savage, on the Ainus, 

Language, the outward expression of 
reason, 193 ; its relation to anthro- 
pology, 193; its physical basis, 193; 
its relation to ethnology, 194; its 
evolution, 195; its origin in a single 
centre, 196; develops species and 
genera which do not mix, 198 9; 
its value as a racial test, 200, 204 ; 
its niorphological orders, 205; their 
evolution, 206; language not com- 
mensurate with thought, note, 193; 
its physical organs, development 
of, 420 

Lanugo, a character of the human 
foetus, and probably of the pre- 
cursor, 175 

Lapps, their ethnical affinities and 
physical characters, 307, 405 

Latham, Dr, his primary human 
groups, 165 

Latin, originally a post-positive lan- 
guage, 214 

Laugerie Basse, human remains found 
at, 148 

Laugerie Haute Cave, its culture eras, 


Lebanon, palaeoliths found in, 93 
Le Conte, J., his theory of evolution, 

note, 49 

Lefevre, A., on the cause of articulate 
speech, 195; on coincidences be- 
tween Old and New World cul- 
tures, 219 
Lemuria, Sclater's, replaced by the 

Indo-African Continent, 229 
Lemuroidea, the, 17 
Lenape Stone, a fraud, 345 
Lenz, O., on the Fans, note, 278 
Leptorrhine nose, its index, 186 
Letourneau, Ch., on the Breton and 
Mauritaninn megaliths, 137 

Lewis, A. L., his theory of the Neo- 
lithic monuments, 137 

Lifynnon Benks Cave, flints found in, 

Linguistic types not stable, 209 10 ; 
evolution, diagram of, 212 

Linne*, his primary human groups, 
164, 222 

Little Falls, Minnesota, rude imple- 
ments found at, 103 

Little Miami river, its flint- bearing 
gravels, 101 

Lockyer, Norman, on the age of the 
Egyptian temples, 56 

Lotherdale Cave, implements found 
in, 8 1 

Louis IX. of France, his reference to 
the "Tartars,'* 304 

Lovisato, Dr D., on Fuegian family 
names, note 4, p. 9 ; on the mental 
capacity of the Fuegians, 48; 6n 
the Fuegian kitchen middens, 96 

Lugard, Capt., on the Gallas, note, 
387 ; on the Wa-Huma of Uganda, 


Lumholtz, C., on the Australian type, 

Lyell, SirCh., his "Seasonal-Migra- 
tion " Theory discussed, 64 66 

MGee, W. J., on palaeolithic man in 
the U. States, 103 

McGwire, J. D., his theory of the Stone 
Ages in N. America, 421 3 

M c Lennan, his theory of the Family, 
8; of the Human Horde, 14 

Madagascar, its ethnical and linguistic 
relations, 205, 285, 330 

Madelenian Age, 87, 88 

Magyars, their ethnical and linguistic 
relations, 309 

Mahn, Dr, on Inflection and Agglu- 
tination, note, 215 

Makololos, their rise and fall, 273 

Malagasy, their ethnical relations, 330 ; 
see also Madagascar 

Malay, meaning of the word, note, 
331; the Malay problem, 328 32 

Malayo-Polynesian speech, explana- 
tion of its wide diffusion, 285 90; 
its linguistic affinities, 331 

Malays, their cerebral structure, 47; 
their craniology, 328; their general 
physical characters, 3 30; their cradle- 
land, 33 1 ; their linguistic relations, 



Mallery, Garrick, his theory of Totem- 
ism, 10; on the mound- builders, 

Man, definition of, 16; his place in 
the order Primates, 17; his rela- 
tions to the Simiidze, 19; resem- 
blances, 25 ; differences, 25, 26 ; 
origin of : creation theory, 2 8 ; 
evolution theory, 29; views of lead- 
ing anthropologists, 31 ; mental 
evolution of, 40; antiquity of, 50; 
Palaeolithic, 71; Neolithic, 108; 
specific unity of, 141 ; varietal di- 
versity of, 1 62 

Man, E. H., on the Andamanese, 256 

Manetta, F., on the early closing of 

the cranial sutures, 44; on the Negro 

temperament, 266 
Maori of N. Zealand, their Oceanic 

affinities, 327, and note, ib. 
Marcilly-sur-Eure, human remains 

found at, 148 

Marocco, evidence of dwarfs in, 247 
Masai, their type and ethnical affinities, 


Mason, O. T., on the evidences of 
palaeolithic man in the U. States, 
103 ; on the peopling of America, 

3 6 5 7 
Mathew, Rev. W., on the origin and 

migrations of the Australians, 291 
Mauritania, its neolithic monuments, 


Max M tiller, on race and language, 

Maxwell, J. R., a Negro writer on 
Negro incapacity, 268 

Mayas, their monuments, 138 

Mayorunas, confused with the Span- 
iards of the Marafton, 339 

Melanesian, twofold meaning of the 
term, 284; types, 282 

Melanesians, their cerebral structure, 
47; their domain, 284; their speech 
more primitive than Malay or Poly- 
nesian, 287 8 

Melanochroi, their type, 228, 379 
80; their range, 370 

" Melano- Papuan" languages, 287 

Menangkabau, cradleland of the his- 
torical Malays, 331 

Menhir, its evolution, 128 9; mean- 
ing of the word, 1 29 ; varieties of, 
129 30; its range, 130 

Mental capacity, determined more by 

the structure than the volume of 
the brain, 44; and physical power 
not correlated, 45 ; and cerebral 
structure always correlated, 46 
Mental growth, arrested by the early 
closing of the cranial sutures, 44, 

Mentone Cave Men, their culture, 

73, .14; their fossil remains, 148 
Mesorrhine nose, its index, 186 
Mestizos of Latin America, a stable 

race, 151, 152 
Mexico, evidences of primitive man 

found in, 100 
Meyer, A. B., on the Kalangsof Java, 

Migrations, early, their relations to 

the megalithic monuments, 136; 

their range during inter-glacial 

times, 231 
Miklukho-Maclay, his comparative 

studies of cerebral structure, 47; 

on craniology as a racial test, 177 ; 

on speech as a racial test, 193 

4 ; on the Negritoes of Malay 

Peninsula, 259; on the Mikrone- 

sians, note, 288 

Mikronesian type and speech, 288 
"Mincopies," see Andamanese 
Mind, to be distinguished from Soul, 

note, 41; a function of brain, 41; 

capable of indefinite expansion, 

47; human, differs in degree only 

from the animal, 48 
Minh-huong half-breeds, note, 154 
Miscegenation, persistence of, from 

pleistocene times, 143, 150 
Missing links, must be postulated, 

note, 57, 235 
Mitla, palace of, 1 38 
Mivart, St George, on the divisions 

Quadrumana and Bimana, 17 ; on 

the origin of Human Reason, 48 
Mixed languages non-existent, 199 
Mixed races, prehistoric and historic, 

150, 151 ; American, 151 ; African, 

153; Asiatic, 154 

Moab, its neolithic monuments, 134 
Mongol, origin and meaning of the 

term, 303 
Mongol race, early migrations and 

interminglings, 297 9; chief sub- 
groups, 299 301 ; earliest records 

of, 301 2 

Mongolia, flints found in, 93 
Mongolic, twofold meaning of the 

term, note, 227 



Mongolo Tatar sub-division, 299; 

objections to this term, 302 3 
Monolithic monuments, 123 
Monolithic nomenclature, 124 
Monosyllabism, not the first but the 

last stage of linguistic growth, 

Montano, Dr, on the affinities of the 

Philippine Islanders, 332 "; 
Monte-Aperto pliocene beds, flints 

found in, 91 
Moorehead, W. K., on the mound - 

builders, 106, 107 

Morais, Polynesian monuments, de- 
scribed, 369 70 

Morgan, his theory of the Family, 8 
Morton, Dr, his Crania Americana, 

Mound-builders, the, their age and 

culture, 106, 107 
Moustier Cave, implements found in, 


Moustierian Age, 86, 87 
Miiller, Fr., his polygenist views, 

157? his Nuba-Fulah Family, 170; 

his "Nuba-Fulah" ethnical family, 

Munro, Dr R., on the Irish and Scotch 

crannogs, 122; on the antiquity of 

man, 140; on his first appearance 

in Europe, 141 

Naga Hills, monoliths in, 130 

Nama Hottentots, 253 

Nasal index, how determined, 185; 
table of, 1 86 

Natal, pakeoliths found in, 93 

Nation, meaning of the term, 14 

Neanderthal skull, not pathological, 
33, 34; its characters,' 145 

Negritoes, normally brachy cephalic, 
178; early records of, 245; former 
wide diffusion of, 246; African, 246 
53 ; Oceanic, 254 63 ; in India, 
evidence for and against, 254 6 ; 
in the Andaman Islands, 256 ; in 
the Malay Peninsula, 257 ; in the 
Philippines, 259; in New Guinea, 
260; in Java, 262; in Timor and 
Flores, note, 262 

Negro race, arrest of its mental growth 
explained, 44; its prognathism, 184; 
its twofold division, 243; see also 
Homo iEthiopicus 

Negroes, African and Oceanic, com- 
parative table of their physical 
characters, 264; their mental quali- 

ties, 265; African unprogressive, 
265; general low state of culture, 
268 j two main sub-divisions, Su- 
danese and Bantu, 269 ; their inter- 
mi ngl ings, 274 ; table of chief 
groups, 275 80 
Neolithic, origin and derivation of the 

word, note, 74 
Neolithic Age, its duration, 55, 116, 


Neolithic craniology in France, 424 
Neolithic culture, table of, i TO 
Neolithic man, 108; various types of, 
149; numerous in W. Europe, 377; 
relations to Homo Caucasicus, 377 

Neolithic monuments, two types of, 
cell and block, 123; their range, 
133; their relations to prehistoric 
migrations, 136; their primary ob- 
ject not worship, but interment, 


New Caledonians, their dentition, 184 
New Grange, its tumulus, 131 
"New Race" in the Nile Valley, its 

ethnical relations, 385 6 
New Zealand, its three Stone Ages, 

95 9 6 
Noetling, Dr F., on pliocene man in 

Burma, 423 4 
North Africa, cradleland of Homo 

Caucasicus, 374 5 
Nose, form of, a racial character, 185; 

correlated to the other features, 185 ; 

various types of, 185, 186 
Nuba-Fulah Family, non-existent, 1 70 
Nuraghi, origin of the, 126 

Oceanic Negritoes, 254 63 

Oceanic Negroes, 280 94 

Ohio Valley, its age, 100; its mounds, 
106, 107 

Oldham, R. D., on the Indo- African 
Continent, note, 229 

Olmo, human remains found at, 148 

Onkilons, their relations to the Chuk- 
chi, 319 


Oppert, Dr G., on the Aryan (Hindu) 

element in India, 398 9 
Orang-utan, 20; range and characters 

of, 20, 22 
Organ and Function, evolution of, 

correlated, 41, 42 
Orohippus, 37 
Orthognathism, how determined, 181 



Osmanli Turks, their type, note, u i ; 

their origin and ethnical relations, 

311 a 
Ostyaks, their physical characters, 


Otomi language, its morphology, 214 
Otta, human (?) implements found at, 

Paccaritambo, 139 
Paddington, eoliths found at, 82 
Palaeolithic, meaning and derivation of 

the word, note, 74 
Palaeolithic Aje, various divisions of, 

8 3 
Palaeolithic cultures, progressive, 73; 

in some places continuous with 

Neolithic, 73 ; grades of, 83 ; table 

of, no, in 
Palaeolithic man, 71 ; spread over the 

globe, 72; materials for his study, 

74; in Indo-China, 423 4 ; see also 

Quaternary man 
Palseolithic races, their remains, 143 

4; ail long-headed, 149 
Palaeoliths, unbulbed, 74 
Palenque, 138 

Palestine, palseolith found in, 93 
Papuan, meaning of the term, note, 

Papuans, normally dolichocephalic, 

178; their domain past and present, 

283 4; their physical characters, 


Papuasia, the Papuan domain, 283 
Paraderos, meaning of the term, note, 

Parker, Theodore, on Negro in- 

capacity, 268 

Pastoral state, see Social states 
Patagonia, evidences of primitive 

man found in, 96, 97 
Patagonian fossil crania of Eskimo 

type, 364 

Paulistas, the, a vigorous mixed race, 

Peat-beds, Danish, a time gauge, 116, 


People, meaning of the term, 14 
Permian Finns, 307 
Persians, their primitive type, 407 


Peruvians, their culture, 139 
Peschel, his primary human groups, 

Petrie, Flinders, his discoveries at 

Coptos, 56 ; in the Abydos district, 

92; his "New Race" in the Nile 
Valley, 385-6 

Petroff, I., on the growth of kitchen 
middens, 77 

Philippine Islanders, their complex 
physical relations, 332 3 

Phonesis, a physical function, 193 4 

Picts, their ethnical relations, 3/89 

Pile-dwellings, see Lake-dwellings 

Pinches, T. G., on Akkadian mono- 
syllabism, 208 

Pitcairn Islanders, vigorous hall- 
breeds, 154 

Pithecanthropus erectus, 24; his phy- 
sical characters, 144 

Placard Cave, its successive culture 
eras, 89, 90 

Platyrrhine nose, its index, 186 

Pleistocene fauna, persistence of, 38; 
man, see Quaternary 

Pliocene man, evidence of, in Argen- 
tina, 98; his physical characters, 
2367; his migrations, 238 

Pliopithecus, 18 

Podbaba, human remains found at, 


Polygenists, their linguistic argument, 


Polylithic monuments, 123 128 ; their 
evolution, 124 

Polylithic nomenclature, 124 

Polynesians, Eastern, a branch of the 
Caucasic family, 288; their speech 
less primitive than the Melanesian, 

Polysynthesis, its nature, 211; differs 
from agglutination, 211; see also 
American Languages 

Pont Newydd Cave, flints found in, 81 

Post-glacial man, 63, 69 

Pouchet, identifies race and species, 5 

Powell, J. W., on the American Lin- 
guistic Families, 360; on the Stone 
Ages in N. America, 421 2 

Predmost, human remains found at, 


Pre-glacial man, 63 
Prehistoric period, its duration, 55 
Prehistoric monuments of America, 

Prestwich, Prof., his objections to 

Croll's periodicity theory, 58 60 
Prichard, Dr, his definition of race, 5; 

on the unity of man, 160; on the 

natural history of man, 165 6 
Primary human groups, various schemes 

of systematists, 1 63 ; the four funda- 



mental, 221 2; their independent 
evolution from a common precursor, 
223; their Family Tree, 224; their 
differentiation, 225; their physical 
and mental characters, 227, 228 

Primates, old and present divisions of 
the, 17, 18 

Prognathism, how determined, 181 

Primer- Bey, Dr, on hair as 2 race 
character, 174 

Psychic argument for the unity of man, 
1 60 

Pueblo Indians, their casas grandes, 


Puy-Courny, implements found at, 91 
Pyramids, Egyptian, '* petrified 

mounds," 128; American and 

Egyptian compared, 369 

Quadrumana, 17 

Quaternary man in Britain, 78 ; in 
France, 82 ; in Africa, 92 ; in Asia, 
93 ; in Australia, 94 ; in New 
Zealand, 95 ; in America, 96 

uecnuas, their culture, 139 
uixeramobim Valley, human remains 
found in, 100 

Race, meaning and value of the term, 
4, 5 ; physical criteria of, 171 ; men- 
tal criteria of, 191 

Racial characters, persistence of, 34 

Raffray, M., on the New Guinea Ne- 
gritoes, 261 

Ray, S. H.,on the "Melano- Papuan" 
languages, 287 8 

Rcade, W., on the Fans, note, 278 

Reclus, Elisee, on the temperament of 
the American aborigines, 353 

Religion, its evolution, 216 7; a 
poor criterion of race, 217 

Rendileh Hamites, 389 

Retzius, A., on the American abori- 
gines, 3467 

Rhys, Prof. J., on the relations of 
Picts and Basques, 378 9 

Rio Carcarana, human remains found 
in, 99 

Rio Negro, Patagonia, implements 
found in, 97 

Rodway, J., on the couvade, 368 

Rolleston, H. B., on heredity, 35 ; 
on an Australian brain, note, 47 

Romanes, Prof, on human and animal 
consciousness, 48 

Romans of N. Africa, fair Berber type 
not due to them, 383 4 

Round Towers, Irish, their probable 
age and origin, 133 

Rudimentary to be distinguished from 
atrophied organs, note 2, p. 26 

Ruffin, Col., on the Negro tempera- 
ment, 267 

St Acheul, flints found at, 82 

St John, Sir Spencer, on the Negro 
temperament, 267 

St Pierre, its alignments, 137 

St Prest, carved bones found at, 91 

Sakai Negritoes, 257 8 

Salmon, Ph., on the ciosing of the 
sutures in fossil crania, 45 ; on the 
craniology of primitive man in 
Europe, 149 50 

Samang Negritoes, 257 8 

Sambaqui, the, of Brazil, roo 

Samborombon, human remains found 
at, 34. 99 

Samoans, their affinities, 326 

Samoyedes, their physical characters 
and affinities, 205 6 

Sanskrit, its tendency to polysyn thesis, 
214; its diffusion in India, 399 

Santa Catharina, Brazil, evidences of 
primitive man found at, roo 

Santarem, Brazil, human remains 
found at, 99 

Sarmatse, of doubtful ethnical affini- 
ties, 399, note, 400 

Sawaiori languages, less primitive than 
the Melanesian, 287 

Sayce, A. H., on the evolution of 
speech, 209, 213 ; on the Blemmyes 
of Nubia, 386 

Schmidt, Prof., his cranial measure- 
ments, 43 

Schrumpf, G. A., on the Armenian 
language, note, 411 

Semites, their relations to the Ha- 
mites, 391; their cradleland, 392; 
their domain and historic divisions, 
392 3 ; their physical and mental 
characters, 393 4 

Semitic languages, their branches and 
structure, 394 5 

Sergi, his Tertiary Horninidae, 32 ; 
rejected, 37 ; on the racial value 
of different physical features, note, 

Sessi, neolithic structures, 126 

Shans of Indo-China, 321 

Shell-mounds, see Kitchen-middens 

Shone, W., on pre-glacial man, 62 

Silbury Hill, a typical barrow, 127 



Simiidse, the, i8j relations of toman, Sweden, its prehistoric monuments, 

Siret, H. and L., on primitive cultures 

in Spain, 127 
Siryanian Finns, 307 
Skull, shape and size of racial tests, 

177, 178 ... 

Slavs, their primitive type, note, 400 
Smith, W. G., on the eoliths of the 

Thames Valley, 82 
Smyth, Brough, on the evidences of 

primitive 'man in Australia, 94 
Social states, pastoral, agricultural, 

&c., not testj of race, 215 
Solons of Kulja and the Amur basin, 

Solutre Cave, implements found in, 


Soiutrian Age, 87 

Somali, their type and ethnical rela- 
tions, 387, 388 
Sommier, S., on the Samoyedes, Ost- 

yaks and Lapps, 206 7 
Soyotes, Siberian Finns, 305, 308 
Species, the physiological test of, 


Specific unity of man, 141 
Speech, see Language 
Spy Cavern, human remains found in, 

Stalagmite, growth of irregular, 76 ; 

hence no sure test of age, 76, 77 
Stanley, H. M., his measurements of 

the Wambutti dwarfs, 189 ; on the 

Wa-Humas, note, 272 
Stature, a racial character, 187 ; table 

of heights, 188 

Stazzone, neolithic structures, 126 
Stem, meaning of the term, 1 1 
Stet Cave, flints found in, 81 
Stock, meaning of the term, 1 1 
Stock languages, no proof of stock 

races, 156, 157 
Stoke Newington, eoliths found at, 


Stokes, W., on Picts and Kelts, 379 
Stone Ages in N. America, J. D. 

McGuire's theory, 421 3 
Stone-circles, 130 
Stonehenge, 131 
Stiibei and Uhle, on the Tiahuanaco 

monuments, 138 9 
Sudan, meaning of the term, note, 

Sudanese Negroes, physical type, 269 ; 

mixed groups, 26970; chief tribes, 

*75~ 7 

Syria, palaeolith found in, 03 
Szeklers, a Magyar people of Tran- 
sylvania, 309 

Table of mixed peoples speaking un- 
mixed languages, 201 ; of peoples 
who^e speech has shifted, 202 ; of 
peoples whose type has changed, 
their speech persisting, 203 ; of the 
Aryan linguistic family, 412 

Tagalog language of .Philippine Is., 
its relations to Gyarung of Central 
Asia, 326 

Tahitians, their affinities, 326 

Talayots, neolithic structures, 126 

Tamahu Libyans of the Egyptian 
records, 384 

Tartar, see Tatar 

Tasmanians, their relations to the 
Australians and to primitive man, 
292 4.; their eolithic culture, 293 

Tatar, origin and meaning of the term, 
303 4 ; earliest records of, 304 

Tatar Khans and Khanates, explana- 
tion of the expressions, 304 

Tavastian Finns, 307 

Teeth, see Dentition 

Ten Kate, Dr H., on the former pre- 
sence of Negritoes in Timor and 
Flores, note, 261 2 

Terramare, prehistoric stations in 
Italy, 404 

Teutonic languages, 413; phonetic 
system, 4134 

Teutons, their ethnical relations in 
Britain, 398; their primitive type, 
397, 408 

Thenay, human (?) implements found 
at, 31, 01 

Thomas, Cyrus, on supposed evidences 
of foreign influences on American 
culture, 3423 

Thought, see Mind 

Thurnam, Dr, on reversion, 35 

Tiahuanaco, its megalithic monuments, 
138 40; their origin, 139 

Tibeto-Chinese sub-group of the 
Mongolic division, 299, 319; their 
range and physical characters, 319 
21 ; their languages, 324 6 

Tierra del Fuego, evidences of primi- 
tive man found in, 96 

Timor, former presence of Negritoes 
in, note, 261 2 

Tlemcen, Algeria, flints found at, 93 



Todas, their ethnical affinities, 418 
Toltecs, their monuments, 138 
Tongans, their Indonesian type, 326, 

Topinard, Dr, his diagram of brain 

weight and volume, 40 ; on mental 

expansion, 48 ; on the hair of the 

white, yellow and black races, 175; 

on craniology, 177; on gn \thism, 

facial and sub-nasal, 182; his table 

of cranial capacity, 1 92 ; on Eskimo 

affinities, 364 
Totemism, meaning and origin of, 10; 

derivation of the word, note, 10 
Tremeirchion Cave, flints found in, 


Trenton, its flint-bearing gravels, 101 
Tribe, evolution of the, 7 ; derivation 

of the word, note, 7, 8 
Trinil, Java, fossil human remains 

found at, 144 
Tristram, Canon, on the Moabite 

monuments, 134 
Tuareg Berbers of fair type, 384 
Tunis', remains of palaeolithic man in, 

"Turanian," derivation of the term, 

note, 302 ; why discredited, ib. 
Turk, see Osmanli 
Turki, true name of the so-called 

" Tatars," 303 ; earliest records of, 

304 ; primitive and later types, 


Turkomans, their Turki affinities, 

Tylor, E. B., on Totemism, 10; on 
the specific unity of man, 155; on 
the influence of Asiatic on Ameri- 
can cultures, 218; on Tasmanian 
culture, 293 

Type, two-fold meaning of the term, 
12; derivation of, note I, p. 12 

Ugrian Finns, 305 

Ugrian Voguls, 306 

Uhle, see StUbel 

Umbrian, a post-positive language, 

United States, evidences of palaeolithic 

man found in, discussed, 100 6 
Upland, Pennsylvania, implements 

found at, 101 

Ural-Altaic linguistic family, 313 4 
Usages, common, a poor test of racial 

affinities, 219 
Uslar, Genl. P. V., on the languages 

of Caucasia, 416 7 

Uxmal, 138 

Uzbegs of the Oxus basin, 312 

Vale of Clwyd Caves, flints found in, 

Valentine, M. S., his North Carolina 

"finds," 342 3 
Vandals of Mauritania, fair Berber 

type not due to them, 383 4 
Van den Gheyn, J., on the Aryan 

cradleland, note, 403 ; on the Gal- 

cha languages, 411 
Van Musschenbroek, on the Kalang 

Negritoes, 262 3 
Varietal diversity of man, 162 
Variety, the physiological test of, 


Vayu language of Nepal, 325 
"Venus de Brassempouy, " 252 
Victoria Cave, Settle, carved bones 

found in, 8 1 
Viracocha, cult of, 139 
Virchow, Prof., on the Otta flints, 

and tertiary man, 31 ; his excessive 

conservatism, 144 
Virey, his primary human groups, 

Volga Finns, 310 i 

Wady Biban el-Moluk, Nile Valley, 

flints found in, 92 
Wa-Huma Hamites, 272; their type 

and ethnical relations, 389 
Waitz, Th., on mixed languages, 199; 

on speech as a basis of classifica- 
tion, 201 
Wallace, A. R., man excluded from 

his theory of evolution, 32, 38; on 

the value of cranial studies, 43 
Walton-on-the-Naze, engraved shell 

found at, 78 

Waruanda, j their type and ethnical 
Watusi, \ relations, 389 
Westermarck, his theory of the 

Family, 8 

' White Allophylians," 305 
Whitney, J. D., on the Calaveras 

skull, 104 
Wilde, Sir W. R., on the Irish 

crannogs, 122 
Wilson, Dr D., on the American 

aborigines, 370 
Wilson, Th., on the flints found in N. 

America, 105, 106; on palceolithic 

man in N. America, 423 
Windisch, Prof., on Picts and Kelts, 


442 INDEX. 

Winkler, Dr H., on Ural-Altaic and Xanthochroi, their type, 228, 379 
Japanese linguistic affinities, 313 80; their range, 379 

Wochua dwarfs, bearded and hairy, Yakuts, East Siberian Turki people, 

note, 177 312 

Wolf, Dr L., on the Batwa Negritoes, Yenisei Finns, 308 

248 Yucatan, its prehistoric cities and 
Wusuns of the Chinese records, 308 monuments, 1 38 

Wyatt, Dr W., on the primitive tribe i 

of Adelaide, Australia, 238 Zulu-Xosa Bantus, 271